Chapter 10

Derivatives of Multivariable
Functions

10.1 Limits
Motivating Questions
• What do we mean by the limit of a function f of two variables at a point
(a, b)?

• What techniques can we use to show that a function of two variables does
not have a limit at a point (a, b)?
• What does it mean for a function f of two variables to be continuous at
a point (a, b)?

In this section, we will study limits of functions of several variables, with

a focus on limits of functions of two variables. In single variable calculus, we
studied the notion of limit, which turned out to be a critical concept that
formed the basis for the derivative and the definite integral. In this section we
will begin to understand how the concept of limit for functions of two variables
is similar to what we encountered for functions of a single variable. The limit
will again be the fundamental idea in multivariable calculus, and we will use
this notion of the limit of a function of several variables to define the important
concept of differentiability later in this chapter. We have already seen its use
in the derivatives of vector-valued functions in Section 9.7.
Let’s begin by reviewing what we mean by the limit of a function of one
variable. We say that a function f has a limit L as x approaches a provided
that we can make the values f (x) as close to L as we like by taking x sufficiently
close (but not equal) to a. We denote this behavior by writing

lim f (x) = L.
x→a

Preview Activity 10.1.1 We investigate the limits of several different func-

tions by working with tables and graphs.
a. Consider the function f defined by

f (x) = 3 − x.

Complete Table 10.1.1.

98
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 99

Table 10.1.1 Values of f (x) = 3 − x.

x −0.2 −0.1 0.0 0.1 0.2

f (x)

What does the table suggest regarding limx→0 f (x)?

b. Explain how your results in (a) are reflected in Figure 10.1.2.

5 y

1
x
-3 -2 -1 1 2 3
-1
Figure 10.1.2 The graph of f (x) = 3 − x.

c. Next, consider
x
g(x) = .
|x|
Complete Table 10.1.3 with values near x = 0, the point at which g is
not defined.
x
Table 10.1.3 Values of g(x) = |x| .

x −0.1 −0.01 −0.001 0.001 0.01 0.1

g(x)

What does this suggest about limx→0 g(x)?

d. Explain how your results in (c) are reflected in Figure 10.1.4.

-1 1

-1

x
Figure 10.1.4 The graph of g(x) = |x| .
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 100

e. Now, let’s examine a function of two variables. Let

f (x, y) = 3 − x − 2y.

Complete Table 10.1.5.

Table 10.1.5 Values of f (x, y) = 3 − x − 2y.

x\y −1.0 −0.1 0.0 0.1 1.0

−1.0 4.2
−0.1 1.1
0.0 2.8
0.1 4.9
1.0 2.0

What does the table suggest about lim(x,y)→(0,0) f (x, y)?

f. Explain how your results in (e) are reflected in Figure 10.1.6. Compare
this limit to the limit in part (a). How are the limits similar and how are
they different?
1.5 y
z
3

1.0
1.5 y

0.5
3

x
x
0.5 1.0 1.5

Figure 10.1.6 Left: The graph of f (x, y) = 3 − x − 2y. Right: A contour

plot.

g. Finally, consider
2xy
g(x, y) = ,
x2 + y 2
which is not defined at (0, 0). Complete Table 10.1.7. Round to three
decinal places.
2xy
Table 10.1.7 Values of g(x, y) = x2 +y 2 .

x\y −1.0 −0.1 0.0 0.1 1.0

−1.0 0.198
−0.1 −0.198
0.0 — 0.000
0.1 −0.198
1.0 0.000

What does this suggest about lim(x,y)→(0,0) g(x, y)?

CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 101

h. Explain how your results are reflected in Figure 10.1.8. Compare this
limit to the limit in part (b). How are the results similar and how are
they different?

z
x
0
-1

x
0
-1
-1 0 y 1

2xy
Figure 10.1.8 Left: The graph of g(x, y) = x2 +y 2 . Right: A contour
plot.

10.1.1 Limits of Functions of Two Variables

In Preview Activity 10.1.1, we recalled the notion of limit from single variable
calculus and saw that a similar concept applies to functions of two variables.
Though we will focus on functions of two variables, for the sake of discussion,
all the ideas we establish here are valid for functions of any number of variables.
In a natural followup to our work in Preview Activity 10.1.1, we now formally
define what it means for a function of two variables to have a limit at a point.
Definition 10.1.9 Given a function f = f (x, y), we say that f has limit L
as (x, y) approaches (a, b) provided that we can make f (x, y) as close to L
as we like by taking (x, y) sufficiently close (but not equal) to (a, b). We write

lim f (x, y) = L.
(x,y)→(a,b)

♦
To investigate the limit of a single variable function, limx→a f (x), we often
consider the behavior of f as x approaches a from the right and from the left.
Similarly, we may investigate limits of two-variable functions, lim(x,y)→(a,b) f (x, y)
by considering the behavior of f as (x, y) approaches (a, b) from various direc-
tions. This situation is more complicated because there are infinitely many
ways in which (x, y) may approach (a, b). In the next activity, we see how it is
important to consider a variety of those paths in investigating whether or not
a limit exists.
Activity 10.1.2 Consider the function f , defined by
y
f (x, y) = p ,
x + y2
2

whose graph is shown below in Figure 10.1.10

CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 102

-1

x
0
-1
-1 0 y 1

y
Figure 10.1.10 The graph of f (x, y) = √ .
x2 +y 2

a. Is f defined at the point (0, 0)? What, if anything, does this say about
whether f has a limit at the point (0, 0)?
b. Values of f (to three decimal places) at several points close to (0, 0) are
shown in Table 10.1.11.
Table 10.1.11 Values of a function f .
x\y −1.000 −0.100 0.000 0.100 1.000
−1.000 −0.707 — 0.000 — 0.707
−0.100 — −0.707 0.000 0.707 —
0.000 −1.000 −1.000 — 1.000 1.000
0.100 — −0.707 0.000 0.707 —
1.000 −0.707 — 0.000 — 0.707

Based on these calculations, state whether f has a limit at (0, 0) and give
an argument supporting your statement. (Hint: The blank spaces in the
table are there to help you see the patterns.)
c. Now we formalize the conjecture from the previous part by considering
what happens if we restrict our attention to different paths. First, we
look at f for points in the domain along the x-axis; that is, we consider
what happens when y = 0. What is the behavior of f (x, 0) as x → 0? If
we approach (0, 0) by moving along the x-axis, what value do we find as
the limit?
d. What is the behavior of f along the line y = x when x > 0; that is, what
is the value of f (x, x) when x > 0? If we approach (0, 0) by moving along
the line y = x in the first quadrant (thus considering f (x, x) as x → 0+ ,
what value do we find as the limit?
e. In general, if lim(x,y)→(0,0) f (x, y) = L, then f (x, y) approaches L as (x, y)
approaches (0, 0), regardless of the path we take in letting (x, y) → (0, 0).
Explain what the last two parts of this activity imply about the existence
of lim(x,y)→(0,0) f (x, y).
f. Shown below in Figure 10.1.12 is a set of contour lines of the function
f . What is the behavior of f (x, y) as (x, y) approaches (0, 0) along any
straight line? How does this observation reinforce your conclusion about
the existence of lim(x,y)→(0,0) f (x, y) from the previous part of this activ-
ity? (Hint: Use the fact that a non-vertical line has equation y = mx for
some constant m.)
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 103

y
Figure 10.1.12 Contour lines of f (x, y) = √ .
x2 +y 2

As we have seen in Activity 10.1.2, if (x, y) approachs (a, b) along two

different paths and we find that f (x, y) has two different limits, we can conclude
that lim(x,y)→(a,b) f (x, y) does not exist. This is similar to the one-variable
example g(x) = x/|x| as shown in Figure 10.1.13; limx→0 g(x) does not exist
because we see different limits as x approaches 0 from the left and the right.

-1 1

-1

x
Figure 10.1.13 The graph of g(x) = |x| .

As a general rule, we have

Limits along different paths.

If f (x, y) has two different limits as (x, y) approaches (a, b) along two
different paths, then lim(x,y)→(a,b) f (x, y) does not exist.

As the next activity shows, studying the limit of a two-variable function f

by considering the behavior of f along various paths can require subtle insights.
Activity 10.1.3 Let’s consider the function g defined by

x2 y
g(x, y) =
x4 + y 2

and investigate the limit lim(x,y)→(0,0) g(x, y).

a. What is the behavior of g on the x-axis? That is, what is g(x, 0) and
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 104

what is the limit of g as (x, y) approaches (0, 0) along the x-axis?

b. What is the behavior of g on the y-axis? That is, what is g(0, y) and
what is the limit of g as (x, y) approaches (0, 0) along the y-axis?
c. What is the behavior of g on the line y = mx? That is, what is g(x, mx)
and what is the limit of g as (x, y) approaches (0, 0) along the line y =
mx?
d. Based on what you have seen so far, do you think lim(x,y)→(0,0) g(x, y)
exists? If so, what do you think its value is?
e. Now consider the behavior of g on the parabola y = x2 ? What is g(x, x2 )
and what is the limit of g as (x, y) approaches (0, 0) along this parabola?
f. State whether the limit lim(x,y)→(0,0) g(x, y) exists or not and provide a
justification of your statement.
This activity shows that we need to be careful when studying the limit of
a two-variable functions by considering its behavior along different paths. If
we find two different paths that result in two different limits, then we may
conclude that the limit does not exist. However, we can never conclude that
the limit of a function exists only by considering its behavior along different
paths.
Generally speaking, concluding that a limit lim(x,y)→(a,b) f (x, y) exists re-
quires a more careful argument.
Example 10.1.14 Consider the function f defined by
x2 y 2
f (x, y) = .
x2 + y 2
We want to know whether lim(x,y)→(0,0) f (x, y) exists.
Note that if either x or y is 0, then f (x, y) = 0. Therefore, if f has a limit
at (0, 0), it must be 0. We will therefore argue that

lim f (x, y) = 0,
(x,y)→(0,0)

by showing that we can make f (x, y) as close to 0 as we wish by taking (x, y)

sufficiently close (but not equal) to (0, 0). In what follows, we view x and y as
being real numbers that are close, but not equal, to 0.
Since 0 ≤ x2 , we have
y 2 ≤ x2 + y 2 ,
which implies that
y2
≤ 1.
x2 + y 2
Multiplying both sides by x2 and observing that f (x, y) ≥ 0 for all (x, y)
gives
x2 y 2 y2
 
2
0 ≤ f (x, y) = 2 = x ≤ x2 .
x + y2 x2 + y 2
Thus, 0 ≤ f (x, y) ≤ x2 . Since x2 → 0 as x → 0, we can make f (x, y) as
close to 0 as we like by taking x sufficiently close to 0 (for this example, it turns
out that we don’t even need to worry about making y close to 0). Therefore,
x2 y 2
lim = 0.
(x,y)→(0,0) x2 + y 2


CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 105

In spite of the fact that these two most recent examples illustrate some
of the complications that arise when studying limits of two-variable functions,
many of the properties that are familiar from our study of single variable
functions hold in precisely the same way.

Properties of Limits.

Let f = f (x, y) and g = g(x, y) be functions so that lim(x,y)→(a,b) f (x, y)

and lim(x,y)→(a,b) g(x, y) both exist. Then
1. lim x = a and lim y=b
(x,y)→(a,b) (x,y)→(a,b)
 
2. lim cf (x, y) = c lim f (x, y) for any scalar c
(x,y)→(a,b) (x,y)→(a,b)

3. lim [f (x, y) ± g(x, y)] = lim f (x, y) ± lim g(x, y)

(x,y)→(a,b) (x,y)→(a,b) (x,y)→(a,b)
  
4. lim [f (x, y)g(x, y)] = lim f (x, y) lim g(x, y)
(x,y)→(a,b) (x,y)→(a,b) (x,y)→(a,b)

lim f (x, y)
f (x, y) (x,y)→(a,b)
5. lim = if lim g(x, y) 6= 0.
(x,y)→(a,b) g(x, y) lim g(x, y) (x,y)→(a,b)
(x,y)→(a,b)

We can use these properties and results from single variable calculus to
verify that many limits exist. For example, these properties show that the
function f defined by

f (x, y) = 3x2 y 3 + 2xy 2 − 3x + 1

has a limit at every point (a, b) and, moreover,

lim f (x, y) = f (a, b).

(x,y)→(a,b)

The reason for this is that polynomial functions of a single variable have
limits at every point.

10.1.2 Continuity
Recall that a function f of a single variable x is said to be continuous at x = a
provided that the following three conditions are satisfied:
1. f (a) exists,
2. limx→a f (x) exists, and
3. limx→a f (x) = f (a).
Using our understanding of limits of multivariable functions, we can define
continuity in the same way.
Definition 10.1.15 A function f = f (x, y) is continuous at the point (a, b)
provided that
1. f is defined at the point (a, b),
2. lim(x,y)→(a,b) f (x, y) exists, and

3. lim(x,y)→(a,b) f (x, y) = f (a, b).

CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 106

♦
For instance, we have seen that the function f defined by f (x, y) = 3x2 y 3 +
2xy 2 − 3x + 1 is continous at every point. And just as with single variable
functions, continuity has certain properties that are based on the properties of
limits.
Properties of continuity.

Let f and g be functions of two variables that are continuous at the

point (a, b). Then
1. cf is continuous at (a, b) for any scalar c
2. f + g is continuous at (a, b)

3. f − g is continuous at (a, b)
4. f g is continuous at (a, b)
f
5. g is continuous at (a, b) if g(a, b) 6= 0

Using these properties, we can apply results from single variable calculus to
decide about continuity of multivariable functions. For example, the coordinate
functions f and g defined by f (x, y) = x and g(x, y) = y are continuous
at every point. We can then use properties of continuity listed to conclude
that every polynomial function in x and y is continuous at every point. For
example, g(x, y) = x2 and h(x, y) = y 3 are continuous functions, so their
product f (x, y) = x2 y 3 is a continuous multivariable function.

10.1.3 Summary
• A function f = f (x, y) has a limit L at a point (a, b) provided that we
can make f (x, y) as close to L as we like by taking (x, y) sufficiently close
(but not equal) to (a, b).
• If (x, y) has two different limits as (x, y) approaches (a, b) along two
different paths, we can conclude that lim(x,y)→(a,b) f (x, y) does not exist.
• Properties similar to those for one-variable functions allow us to conclude
that many limits exist and to evaluate them.
• A function f = f (x, y) is continuous at a point (a, b) in its domain if f
has a limit at (a, b) and
f (a, b) = lim f (x, y).
(x,y)→(a,b)

10.1.4 Exercises
1. Find the limits, if they exist, or type DNE for any which do not exist.

1x2
lim
(x,y)→(0,0) 3x2 + 4y 2

1) Along the x-axis:

2) Along the y-axis:
3) Along the line y = mx :
4) The limit is:
Answer 1. 0.333333333333333
Answer 2. 0
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 107

1
Answer 3. 3+4m2
Answer 4. dne
2. Determining the limit of a funtion. In this problem we show that
the function
6x − y
f (x, y) =
x+y
does not have a limit as (x, y) → (0, 0).
(a) Suppose that we consider (x, y) → (0, 0) along the curve y = 4x.
Find the limit in this case:
6x−y
lim x+y =
(x,4x)→(0,0)
(b) Now consider (x, y) → (0, 0) along the curve y = 5x. Find the
limit in this case:
6x−y
lim x+y =
(x,5x)→(0,0)
(c) Note that the results from (a) and (b) indicate that f has no limit
as (x, y) → (0, 0) (be sure you can explain why!).
To show this more generally, consider (x, y) → (0, 0) along the curve
y = mx, for arbitrary m. Find the limit in this case:
6x−y
lim x+y =
(x,mx)→(0,0)
(Be sure that you can explain how this result also indicates that f has
no limit as (x, y) → (0, 0).
6−41
Answer 1. 1+41
6−51
Answer 2. 1+51
6−m
Answer 3. 1+m
3. Show that the function
x3 y
f (x, y) = .
x6+ y3

does not have a limit at (0, 0) by examining the following limits.

(a) Find the limit of f as (x, y) → (0, 0) along the line y = x.
lim f (x, y) =
(x,y)→(0,0)
y=x

(b) Find the limit of f as (x, y) → (0, 0) along the curve y = x3 .

lim f (x, y) =
(x,y)→(0,0)
y=x3

(Be sure that you are able to explain why the results in (a) and (b)
indicate that f does not have a limit at (0,0)!
Answer 1. 0
Answer 2. 1
4. Find the limit, if it exists, or type N if it does not exist.
3x2
lim =
(x,y)→(0,0) 3x + 5y 2
2

Answer. N
5. Find the limit, if it exists, or type N if it does not exist.
(x + 13y)2
lim 2 =
(x,y)→(0,0) x2 + 13 y 2
Answer. N
6. Find the limit,√if it exists, or type ’DNE’ if it does not exist.
2 2
lim e 4x +3y =
(x,y)→(1,3)
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 108

Answer. 261.848047157352
7. Find the limit, if it exists, or type N if it does not exist.
4xy + 2yz + 2xz
lim =
(x,y,z)→(0,0,0) 16x2 + 4y 2 + 4z 2
Answer. N
8. Find the limit, if it exists, or type N if it does not exist.
2 2
2zex +y
lim =
(x,y,z)→(5,1,4) 5x2 + y 2 + 4z 2
Answer. 8241246712.79321
9. Find the limit (enter ’DNE’ if the limit does not exist)
Hint: rationalize the denominator.
(9x2 + 2y 2 )
lim p
(x,y)→(0,0) (9x2 + 2y 2 + 1) − 1

Answer. 2
10. The largest set on which the function f (x, y) = 1/(3−x2 −y 2 ) is continuous
is
A. All of the xy-plane
B. The interior of the circle x2 + y 2 = 3
C. The exterior of the circle x2 + y 2 = 3

D. The interior of the circle x2 + y 2 = 3, plus the circle

E. All of the xy-plane except the circle x2 + y 2 = 3

Answer. E
xy
11. Consider the function f defined by f (x, y) = x2 +y 2 +1 .
a. What is the domain of f ?

b. Evaluate limit of f at (0, 0) along the following paths: x = 0, y = 0,

y = x, and y = x2 .
c. What do you conjecture is the value of lim(x,y)→(0,0) f (x, y)?

d. Is f continuous at (0, 0)? Why or why not?

e. Use appropriate technology to sketch both surface and contour plots
of f near (0, 0). Write several sentences to say how your plots affirm
your findings in (a) - (d).
xy
12. Consider the function g defined by g(x, y) = x2 +y 2 .
a. What is the domain of g?

b. Evaluate limit of g at (0, 0) along the following paths: x = 0, y = x,

and y = 2x.
c. What can you now say about the value of lim(x,y)→(0,0) g(x, y)?
d. Is g continuous at (0, 0)? Why or why not?

e. Use appropriate technology to sketch both surface and contour plots

of g near (0, 0). Write several sentences to say how your plots affirm
your findings in (a) - (d).
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 109

2x2 y
13. Consider the function h defined by h(x, y) = x4 +y 2 .
a. What is the domain of h?
b. Evaluate the limit of h at (0, 0) along all linear paths the contain the
origin. What does this tell us about lim(x,y)→(0,0) h(x, y)? (Hint: A
non-vertical line throught the origin has the form y = mx for some
constant m.)
c. Does lim(x,y)→(0,0) h(x, y) exist? Verify your answer. Check by using
appropriate technology to sketch both surface and contour plots of
h near (0, 0). Write several sentences to say how your plots affirm
your findings about lim(x,y)→(0,0) h(x, y).
14. For each of the following prompts, provide an example of a function of
two variables with the desired properties (with justification), or explain
why such a function does not exist.
a. A function p that is defined at (0, 0), but lim(x,y)→(0,0) p(x, y) does
not exist.
b. A function q that does not have a limit at (0, 0), but that has the
same limiting value along any line y = mx as x → 0.
c. A function r that is continuous at (0, 0), but lim(x,y)→(0,0) r(x, y)
does not exist.
d. A function s such that

lim s(x, x) = 3 and lim s(x, 2x) = 6,

(x,x)→(0,0) (x,2x)→(0,0)

for which lim(x,y)→(0,0) s(x, y) exists.

e. A function t that is not defined at (1, 1) but lim(x,y)→(1,1) t(x, y)
does exist.
15. Use the properties of continuity to determine the set of points at which
each of the following functions is continuous. Justify your answers.
x+2y
a. The function f defined by f (x, y) = x−y

sin(x)
b. The function g defined by g(x, y) = 1+ey

c. The function h defined by

(
xy
x2 +y 2 if (x, y) 6= (0, 0)
h(x, y) =
0 if (x, y) = (0, 0)

d. The function k defined by

x2 y 4
(
x2 +y 2 if (x, y) 6= (0, 0)
k(x, y) =
0 if (x, y) = (0, 0)
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 110

10.2 First-Order Partial Derivatives

Motivating Questions
• How are the first-order partial derivatives of a function f of the indepen-
dent variables x and y defined?
• Given a function f of the independent variables x and y, what do the
first-order partial derivatives ∂f ∂f
∂x and ∂y tell us about f ?

The derivative plays a central role in first semester calculus because it

provides important information about a function. Thinking graphically, for
instance, the derivative at a point tells us the slope of the tangent line to the
graph at that point. In addition, the derivative at a point also provides the
instantaneous rate of change of the function with respect to changes in the
independent variable.
Now that we are investigating functions of two or more variables, we can
still ask how fast the function is changing, though we have to be careful about
what we mean. Thinking graphically again, we can try to measure how steep
the graph of the function is in a particular direction. Alternatively, we may
want to know how fast a function’s output changes in response to a change
in one of the inputs. Over the next few sections, we will develop tools for
addressing issues such as these these. Preview Activity 10.2.1 explores some
issues with what we will come to call partial derivatives.
Preview Activity 10.2.1 Suppose we take out an $18,000 car loan at interest
rate r and we agree to pay off the loan in t years. The monthly payment, in
dollars, is
1500r
M (r, t) = .
r −12t


1 − 1 + 12

a. What is the monthly payment if the interest rate is 3% so that r = 0.03,

and we pay the loan off in t = 4 years?
b. Suppose the interest rate is fixed at 3%. Express M as a function f of t
alone using r = 0.03. That is, let f (t) = M (0.03, t). Sketch the graph of
f on the left of Figure 10.2.1. Explain the meaning of the function f .
1000 1000
f (t) g(r)

750 750

500 500

250 250

t r
1 2 3 4 5 6 7 8 9 10 0.02 0.04 0.06 0.08 0.10

Figure 10.2.1 Left: Graphs f (t) = M (0.03, t). Right: Graph g(r) =
M (r, 4).

c. Find the instantaneous rate of change f 0 (4) and state the units on this
quantity. What information does f 0 (4) tell us about our car loan? What
information does f 0 (4) tell us about the graph you sketched in (b)?
d. Express M as a function of r alone, using a fixed time of t = 4. That is,
let g(r) = M (r, 4). Sketch the graph of g on the right of Figure 10.2.1.
Explain the meaning of the function g.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 111

e. Find the instantaneous rate of change g 0 (0.03) and state the units on
this quantity. What information does g 0 (0.03) tell us about our car loan?
What information does g 0 (0.03) tell us about the graph you sketched in
(d)?

10.2.1 First-Order Partial Derivatives

In Section 9.1, we studied the behavior of a function of two or more variables
by considering the traces of the function. Recall that in one example, we
considered the function f defined by

x2 sin(2y)
f (x, y) = ,
32
which measures the range, or horizontal distance, in feet, traveled by a projec-
tile launched with an initial speed of x feet per second at an angle y radians
to the horizontal. The graph of this function is given again on the left in
Figure 10.2.2. Moreover, if we fix the angle y = 0.6, we may view the trace
f (x, 0.6) as a function of x alone, as seen at right in Figure 10.2.2.
z 1000
f (x, 0.6)
1500

1000 800
500
600
x
200
400
y 150
1.5 100
1.0
0.5
50 200
0
x
50 100 150 200
x2 sin(2y)
Figure 10.2.2 Left: The trace of z = 32 with y = 0.6.
Since the trace is a one-variable function, we may consider its derivative
just as we did in the first semester of calculus. With y = 0.6, we have

sin(1.2) 2
f (x, 0.6) = x ,
32
and therefore
d sin(1.2)
[f (x, 0.6)] = x.
dx 16
When x = 150, this gives

d sin(1.2)
[f (x, 0.6)]|x=150 = 150 ≈ 8.74 feet per feet per second,
dx 16
which gives the slope of the tangent line shown on the right of Figure 10.2.2.
Thinking of this derivative as an instantaneous rate of change implies that if
we increase the initial speed of the projectile by one foot per second, we expect
the horizontal distance traveled to increase by approximately 8.74 feet if we
hold the launch angle constant at 0.6 radians.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 112

By holding y fixed and differentiating with respect to x, we obtain the first-

order partial derivative of f with respect to x. Denoting this partial derivative
as fx , we have seen that
d f (150 + h, 0.6) − f (150, 0.6)
fx (150, 0.6) = f (x, 0.6)|x=150 = lim .
dx h→0 h
More generally, we have
f (a + h, b) − f (a, b)
fx (a, b) = lim ,
h→0 h
provided this limit exists.
In the same way, we may obtain a trace by setting, say, x = 150 as shown
in Figure 10.2.3.
z 1000
f (150, y)
1500

1000 800
500
600
x
200
400
y 150
1.5 100
1.0
0.5
50 200
0
y
0.25 0.50 0.75 1.00 1.25
x2 sin(2y)
Figure 10.2.3 The trace of z = 32 with x = 150.
This gives
1502
f (150, y) = sin(2y),
32
and therefore
d 1502
[f (150, y)] = cos(2y).
dy 16
If we evaluate this quantity at y = 0.6, we have
d 1502
[f (150, y)]|y=0.6 = cos(1.2) ≈ 509.5 feet per radian.
dy 16
Once again, the derivative gives the slope of the tangent line shown on the
right in Figure 10.2.3. Thinking of the derivative as an instantaneous rate of
change, we expect that the range of the projectile increases by 509.5 feet for
every radian we increase the launch angle y if we keep the initial speed of the
projectile constant at 150 feet per second.
By holding x fixed and differentiating with respect to y, we obtain the first-
order partial derivative of f with respect to y. As before, we denote this partial
derivative as fy and write
d f (150, 0.6 + h) − f (150, 0.6)
fy (150, 0.6) = f (150, y)|y=0.6 = lim .
dy h→0 h
As with the partial derivative with respect to x, we may express this quan-
tity more generally at an arbitrary point (a, b). To recap, we have now arrived
at the formal definition of the first-order partial derivatives of a function of
two variables.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 113

Definition 10.2.4 The first-order partial derivatives of f with respect

to x and y at a point (a, b) are, respectively,

f (a + h, b) − f (a, b)
fx (a, b) = lim , and
h→0 h
f (a, b + h) − f (a, b)
fy (a, b) = lim ,
h→0 h
provided the limits exist. ♦
Activity 10.2.2 Consider the function f defined by

xy 2
f (x, y) =
x+1
at the point (1, 2).
a. Write the trace f (x, 2) at the fixed value y = 2. On the left side of
Figure 10.2.5, draw the graph of the trace with y = 2 around the point
where x = 1, indicating the scale and labels on the axes. Also, sketch
the tangent line at the point x = 1.

xy 2
Figure 10.2.5 Traces of f (x, y) = x+1 .

b. Find the partial derivative fx (1, 2) and relate its value to the sketch you
just made.

c. Write the trace f (1, y) at the fixed value x = 1. On the right side of
Figure 10.2.5, draw the graph of the trace with x = 1 indicating the scale
and labels on the axes. Also, sketch the tangent line at the point y = 2.
d. Find the partial derivative fy (1, 2) and relate its value to the sketch you
just made.
As these examples show, each partial derivative at a point arises as the
derivative of a one-variable function defined by fixing one of the coordinates. In
addition, we may consider each partial derivative as defining a new function of
the point (x, y), just as the derivative f 0 (x) defines a new function of x in single-
variable calculus. Due to the connection between one-variable derivatives and
partial derivatives, we will often use Leibniz-style notation to denote partial
derivatives by writing
∂f ∂f
(a, b) = fx (a, b), and (a, b) = fy (a, b).
∂x ∂y
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 114

To calculate the partial derivative fx , we hold y fixed and thus we treat y

as a constant. In Leibniz notation, observe that
∂ ∂
(x) = 1 and (y) = 0.
∂x ∂x
To see the contrast between how we calculate single variable derivatives and
d ∂
partial derivatives, and the difference between the notations dx [ ] and ∂x [ ],
observe that
d d d d
[3x2 − 2x + 3] = 3 [x2 ] − 2 [x] + [3] = 3 · 2x − 2,
dx dx dx dx
∂ 2 ∂ ∂ ∂
and [x y − xy + 2y] = y [x2 ] − y [x] + [2y] = y · 2x − y
∂x ∂x ∂x ∂x
Thus, computing partial derivatives is straightforward: we use the standard
rules of single variable calculus, but do so while holding one (or more) of the
variables constant.
Activity 10.2.3
a. If f (x, y) = 3x3 − 2x2 y 5 , find the partial derivatives fx and fy .

xy 2
b. If f (x, y) = , find the partial derivatives fx and fy .
x+1
c. If g(r, s) = rs cos(r), find the partial derivatives gr and gs .
d. Assuming f (w, x, y) = (6w + 1) cos(3x2 + 4xy 3 + y), find the partial
derivatives fw , fx , and fy .

x2t z 3
e. Find all possible first-order partial derivatives of q(x, t, z) = .
1 + x2

10.2.2 Interpretations of First-Order Partial Derivatives

Recall that the derivative of a single variable function has a geometric interpre-
tation as the slope of the line tangent to the graph at a given point. Similarly,
we have seen that the partial derivatives measure the slope of a line tangent
to a trace of a function of two variables as shown in Figure 10.2.6.
z z
1500 1500

1000 1000

500 500

x x
200 200
y 150 y 150
1.5 100 1.5 100
1.0 50 1.0 50
0.5 0.5
0 0

Figure 10.2.6 Tangent lines to two traces of the distance function.

Now we consider the first-order partial derivatives in context. Recall that
the difference quotient f (a+h)−f
h
(a)
for a function f of a single variable x at a
point where x = a tells us the average rate of change of f over the interval
[a, a + h], while the derivative f 0 (a) tells us the instantaneous rate of change
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 115

of f at x = a. We can use these same concepts to explain the meanings of the

partial derivatives in context.
Activity 10.2.4 The speed of sound C traveling through ocean water is a
function of temperature, salinity and depth. It may be modeled by the function

C = 1449.2 + 4.6T − 0.055T 2 + 0.00029T 3 + (1.34 − 0.01T )(S − 35) + 0.016D.

Here C is the speed of sound in meters/second, T is the temperature in

degrees Celsius, S is the salinity in grams/liter of water, and D is the depth
below the ocean surface in meters.

a. State the units in which each of the partial derivatives, CT , CS and CD ,

are expressed and explain the physical meaning of each.
b. Find the partial derivatives CT , CS and CD .
c. Evaluate each of the three partial derivatives at the point where T = 10,
S = 35 and D = 100. What does the sign of each partial derivatives tell
us about the behavior of the function C at the point (10, 35, 100)?

10.2.3 Using tables and contours to estimate partial deriva-

tives
Remember that functions of two variables are often represented as either a
table of data or a contour plot. In single variable calculus, we saw how we can
use the difference quotient to approximate derivatives if, instead of an algebraic
formula, we only know the value of the function at a few points. The same
idea applies to partial derivatives.
Activity 10.2.5 The wind chill, as frequently reported, is a measure of how
cold it feels outside when the wind is blowing. In Table 10.2.7, the wind chill
w, measured in degrees Fahrenheit, is a function of the wind speed v, measured
in miles per hour, and the ambient air temperature T , also measured in degrees
Fahrenheit. We thus view w as being of the form w = w(v, T ).
Table 10.2.7 Wind chill as a function of wind speed and temperature.
v\T −30 −25 −20 −15 −10 −5 0 5 10 15 20
5 −46 −40 −34 −28 −22 −16 −11 −5 1 7 13
10 −53 −47 −41 −35 −28 −22 −16 −10 −4 3 9
15 −58 −51 −45 −39 −32 −26 −19 −13 −7 0 6
20 −61 −55 −48 −42 −35 −29 −22 −15 −9 −2 4
25 −64 −58 −51 −44 −37 −31 −24 −17 −11 −4 3
30 −67 −60 −53 −46 −39 −33 −26 −19 −12 −5 1
35 −69 −62 −55 −48 −41 −34 −27 −21 −14 −7 0
40 −71 −64 −57 −50 −43 −36 −29 −22 −15 −8 −1

a. Estimate the partial derivative wv (20, −10). What are the units on this
quantity and what does it mean? (Recall that we can estimate a partial
derivative of a single variable function f using the symmetric difference
quotient f (x+h)−f
2h
(x−h)
for small values of h. A partial derivative is a
derivative of an appropriate trace.)
b. Estimate the partial derivative wT (20, −10). What are the units on this
quantity and what does it mean?
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 116

c. Use your results to estimate the wind chill w(18, −10). (Recall from single
variable calculus that for a function f of x, f (x + h) ≈ f (x) + hf 0 (x).)

d. Use your results to estimate the wind chill w(20, −12).

e. Consider how you might combine your previous results to estimate the
wind chill w(18, −12). Explain your process.
Activity 10.2.6 Shown below in Figure 10.2.8 is a contour plot of a function
f . The values of the function on a few of the contours are indicated to the left
of the figure.
3 y

2
-1

1
0
x
1
-3 -2 -1 1 2 3
2
3 -1
4
5 -2
6
-3

Figure 10.2.8 A contour plot of f .

a. Estimate the partial derivative fx (−2, −1). (Hint: How can you find
values of f that are of the form f (−2 + h) and f (−2 − h) so that you can
use a symmetric difference quotient?)

b. Estimate the partial derivative fy (−2, −1).

c. Estimate the partial derivatives fx (−1, 2) and fy (−1, 2).
d. Locate, if possible, one point (x, y) where fx (x, y) = 0.

e. Locate, if possible, one point (x, y) where fx (x, y) < 0.

f. Locate, if possible, one point (x, y) where fy (x, y) > 0.
g. Suppose you have a different function g, and you know that g(2, 2) = 4,
gx (2, 2) > 0, and gy (2, 2) > 0. Using this information, sketch a possibility
for the contour g(x, y) = 4 passing through (2, 2) on the left side of
Figure 10.2.9. Then include possible contours g(x, y) = 3 and g(x, y) = 5.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 117

4 y 4 y

3 3

2 2

1 1

x x
1 2 3 4 1 2 3 4
Figure 10.2.9 Plots for contours of g and h.

h. Suppose you have yet another function h, and you know that h(2, 2) = 4,
hx (2, 2) < 0, and hy (2, 2) > 0. Using this information, sketch a possi-
ble contour h(x, y) = 4 passing through (2, 2) on the right side of Fig-
ure 10.2.9. Then include possible contours h(x, y) = 3 and h(x, y) = 5.

10.2.4 Summary
• If f = f (x, y) is a function of two variables, there are two first order
partial derivatives of f : the partial derivative of f with respect to x,

∂f f (x + h, y) − f (x, y)
(x, y) = fx (x, y) = lim ,
∂x h→0 h
and the partial derivative of f with respect to y,

∂f f (x, y + h) − f (x, y)
(x, y) = fy (x, y) = lim ,
∂y h→0 h

where each partial derivative exists only at those points (x, y) for which
the limit exists.
• The partial derivative fx (a, b) tells us the instantaneous rate of change of
f with respect to x at (x, y) = (a, b) when y is fixed at b. Geometrically,
the partial derivative fx (a, b) tells us the slope of the line tangent to the
y = b trace of the function f at the point (a, b, f (a, b)).
• The partial derivative fy (a, b) tells us the instantaneous rate of change of
f with respect to y at (x, y) = (a, b) when x is fixed at a. Geometrically,
the partial derivative fy (a, b) tells us the slope of the line tangent to the
x = a trace of the function f at the point (a, b, f (a, b)).

10.2.5 Exercises
1. Find the first partial derivatives of
x − 4y
f (x, y) = at the point (x, y) = (4, 1).
x + 4y
∂f
(4, 1) =
∂x
∂f
(4, 1) =
∂y
Answer 1. 0.125
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 118

Answer 2. −0.5
2. Find the first partial derivatives of f (x, y) = sin(x − y) at the point (-4,
-4).
A. fx (−4, −4) =
B. fy (−4, −4) =
Answer 1. 1
Answer 2. −1
3. Find the partial derivatives of the function
p
w = 4r2 + 3s2 + 7t2
∂w
∂r =
∂w
∂s =
∂w
∂t =
4r
Answer 1. √
4r 2 +3s2 +7t2
3s
Answer 2. √
4r 2 +3s2 +7t2
7t
Answer 3. √
4r 2 +3s2 +7t2

4. Suppose that f (x, y) is a smooth function and that its partial derivatives
have the values, fx (0, 9) = −4 and fy (0, 9) = −2. Given that f (0, 9) =
1, use this information to estimate the value of f (1, 10). Note this is
analogous to finding the tangent line approximation to a function of one
variable. In fancy terms, it is the first Taylor approximation.
Estimate of (integer value) f (0, 10)
Estimate of (integer value) f (1, 9)
Estimate of (integer value) f (1, 10)
Answer 1. −1
Answer 2. −3
Answer 3. −5
5. The gas law for a fixed mass m of an ideal gas at absolute temperature T ,
pressure P , and volume V is P V = mRT , where R is the gas constant.
Find the partial derivatives
∂P
=
∂V
∂V
=
∂T
∂T
=
∂P
∂P ∂V ∂T
= (an integer)
∂V ∂T ∂P
Answer 1. −mRT V2
mR
Answer 2. P
V
Answer 3. mR
Answer 4. −1
6. Find the first partial derivatives of f (x, y, z) = z arctan( xy ) at the point
(4, 4, -3).
A. ∂f
∂x (4, 4, −3) =
∂f
B. ∂y (4, 4, −3) =
∂f
C. ∂z (4, 4, −3) =
Answer 1. 0.375
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 119

Answer 2. −0.375
Answer 3. 0.785398163397448
7. Find the partial derivatives of the function
Z x
f (x, y) = cos(7t2 + 8t − 2) dt
y

fx (x, y) =
fy (x, y) =


Answer 1. cos 7x2 + 8x + (−2)



Answer 2. − cos 7y 2 + 8y + (−2)
8. Let f (x, y) = e−2x sin(4y).
(a) Using difference quotients with ∆x = 0.1 and ∆y = 0.1, we esti-
mate
fx (2, −2) ≈
fy (2, −2) ≈
(b) Using difference quotients with ∆x = 0.01 and ∆y = 0.01, we find
better estimates:
fx (2, −2) ≈
fy (2, −2) ≈
Answer 1. 0.0328473
Answer 2. 0.00392661
Answer 3. 0.0358814
Answer 4. −0.0092074
9. Determine the sign of fx and fy at each indicated point using the contour
diagram of f shown below. (The point P is that in the first quadrant, at
a positive x and y value; Q through T are located clockwise from P , so
that Q is at a positive x value and negative y, etc.)

(a) At point P ,
fx is ( positive  negative) and
fy is ( positive  negative) .
(b) At point Q,
fx is ( positive  negative) and
fy is ( positive  negative) .
(c) At point S,
fx is ( positive  negative) and
fy is ( positive  negative) .
Answer 1. negative
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 120

Answer 2. negative
Answer 3. negative
Answer 4. positive
Answer 5. positive
Answer 6. negative
10. Your monthly car payment in dollars is P = f (P0 , t, r), where $P0 is the
amount you borrowed, t is the number of months it takes to pay off the
loan, and r percent is the interest rate.
(a) Is ∂P/∂t positive or negative? ( positive  negative)
Suppose that your bank tells you that the magnitude of ∂P/∂t is 15.
What are the units of this value?
(For this problem, write our your units in full, writing dollars for
$, months for months, percent for %, etc. Note that fractional units
generally have a plural numerator and singular denominator.)
(b) Is ∂P/∂r positive or negative? ( positive  negative)
Suppose that your bank tells you that the magnitude of ∂P/∂r is 25.
What are the units of this value?
(For this problem, write our your units in full, writing dollars for
$, months for months, percent for %, etc. Note that fractional units
generally have a plural numerator and singular denominator.)
For both parts of this problem, be sure you can explain what the prac-
tical meanings of the partial derivatives are.
Answer 1. negative
Answer 2. dollars/month
Answer 3. positive
Answer 4. dollars/percent
11. An experiment to measure the toxicity of formaldehyde yielded the data
in the table below. The values show the percent, P = f (t, c), of rats
surviving an exposure to formaldehyde at a concentration of c (in parts
per million, ppm) after t months.
t = 14 t = 16 t = 18 t = 20 t = 22 t = 24
c=0 100 100 100 99 97 95
c=2 100 99 98 97 95 92
c=6 96 95 93 90 86 80
c = 15 96 93 82 70 58 36
(a) Estimate ft (18, 6):
ft (18, 6) ≈
(b) Estimate fc (18, 6):
fc (18, 6) ≈
(Be sure that you can give the practical meaning of these two values
in terms of formaldehyde toxicity.)
90−95
Answer 1. 20−16
82−98
Answer 2. 15−2
12. An airport can be cleared of fog by heating the air. The amount of heat
required depends on the air temperature and the wetness of the fog. The
figure below shows the heat H(T, w) required (in calories per cubic meter
of fog) as a function of the temperature T (in degrees Celsius) and the
water content w (in grams per cubic meter of fog). Note that this figure
is not a contour diagram, but shows cross-sections of H with w fixed at
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 121

0.1, 0.2, 0.3, and 0.4.

(a) Estimate HT (10, 0.2):

HT (10, 0.2) ≈
(Be sure you can interpret this partial derivative in practical terms.)
(b) Make a table of values for H(T, w) from the figure, and use it to
estimate HT (T, w) for each of the following:
T = 10, w = 0.2 : HT (T, w) ≈
T = 20, w = 0.2 : HT (T, w) ≈
T = 10, w = 0.3 : HT (T, w) ≈
T = 20, w = 0.3 : HT (T, w) ≈
(c) Repeat (b) to find Hw (T, w) for each of the following:
T = 10, w = 0.2 : Hw (T, w) ≈
T = 20, w = 0.2 : Hw (T, w) ≈
T = 10, w = 0.3 : Hw (T, w) ≈
T = 20, w = 0.3 : Hw (T, w) ≈
(Be sure you can interpret this partial derivative in practical terms.)
190−240
Answer 1. 10
190−240
Answer 2. 10
160−190
Answer 3. 10
260−330
Answer 4. 10
220−260
Answer 5. 10
330−240
Answer 6. 0.1
260−190
Answer 7. 0.1
450−330
Answer 8. 0.1
350−260
Answer 9. 0.1
13. The Heat Index, I, (measured in apparent degrees F ) is a function of the
actual temperature T outside (in degrees F) and the relative humidity H
(measured as a percentage). A portion of the table which gives values for
this function, I = I(T, H), is reproduced in Table 10.2.10.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 122

Table 10.2.10 A portion of the wind chill data.

T ↓\H → 70 75 80 85
90 106 109 112 115
92 112 115 119 123
94 118 122 127 132
96 125 130 135 141

a. State the limit definition of the value IT (94, 75). Then, estimate
IT (94, 75), and write one complete sentence that carefully explains
the meaning of this value, including its units.
b. State the limit definition of the value IH (94, 75). Then, estimate
IH (94, 75), and write one complete sentence that carefully explains
the meaning of this value, including its units.
c. Suppose you are given that IT (92, 80) = 3.75 and IH (92, 80) = 0.8.
Estimate the values of I(91, 80) and I(92, 78). Explain how the
partial derivatives are relevant to your thinking.
d. On a certain day, at 1 p.m. the temperature is 92 degrees and the
relative humidity is 85%. At 3 p.m., the temperature is 96 degrees
and the relative humidity 75%. What is the average rate of change
of the heat index over this time period, and what are the units on
your answer? Write a sentence to explain your thinking.
14. Let f (x, y) = 12 xy 2 represent the kinetic energy in Joules of an object of
mass x in kilograms with velocity y in meters per second. Let (a, b) be
the point (4, 5) in the domain of f .
a. Calculate fx (a, b).
b. Explain as best you can in the context of kinetic energy what the
partial derivative

f (a + h, b) − f (a, b)
fx (a, b) = lim
h→0 h
tells us about kinetic energy.

c. Calculate fy (a, b).

d. Explain as best you can in the context of kinetic energy what the
partial derivative

f (a, b + h) − f (a, b)
fy (a, b) = lim
h→0 h
tells us about kinetic energy.
e. Often we are given certain graphical information about a function
instead of a rule. We can use that information to approximate par-
tial derivatives. For example, suppose that we are given a contour
plot of the kinetic energy function (as in Figure 10.2.11) instead of a
formula. Use this contour plot to approximate fx (4, 5) and fy (4, 5)
as best you can. Compare to your calculations from earlier parts of
this exercise.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 123

8 y
7
6
5
70 80
4 50 60
3 40
30
20
2
10
1
x
1 2 3 4 5 6 7 8

Figure 10.2.11 The graph of f (x, y) = 12 xy 2 .

15. The temperature on an unevenly heated metal plate positioned in the first
quadrant of the xy-plane is given by
25xy + 25
C(x, y) = .
(x − 1)2 + (y − 1)2 + 1

Assume that temperature is measured in degrees Celsius and that x

and y are each measured in inches. (Note: At no point in the following
questions should you expand the denominator of C(x, y).)
∂C ∂C
a. Determine ∂x |(x,y) and ∂y |(x,y) .

b. If an ant is on the metal plate, standing at the point (2, 3), and starts
walking in the direction parallel to the positive y axis, at what rate
will the temperature the ant is experiencing change? Explain, and
include appropriate units.
c. If an ant is walking along the line y = 3 in the positive x direction, at
what instantaneous rate will the temperature the ant is experiencing
change when the ant passes the point (1, 3)?
d. Now suppose the ant is stationed at the point (6, 3) and walks in a
straight line towards the point (2, 0). Determine the average rate of
change in temperature (per unit distance traveled) the ant encoun-
ters in moving between these two points. Explain your reasoning
carefully. What are the units on your answer?
16. Consider the function f defined by f (x, y) = 8 − x2 − 3y 2 .
a. Determine fx (x, y) and fy (x, y).
b. Find parametric equations in R3 for the tangent line to the trace
f (x, 1) at x = 2.

c. Find parametric equations in R3 for the tangent line to the trace

f (2, y) at y = 1.
d. State respective direction vectors for the two lines determined in (b)
and (c).

e. Determine the equation of the plane that passes through the point
(2, 1, f (2, 1)) whose normal vector is orthogonal to the direction vec-
tors of the two lines found in (b) and (c).
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 124

f. Use a graphing utility to plot both the surface z = 8 − x2 − 3y 2 and

the plane from (e) near the point (2, 1). What is the relationship
between the surface and the plane?
17. Recall from single variable calculus that, given the derivative of a single
variable function and an initial condition, we can integrate to find the
original function. We can sometimes use the same process for functions of
more than one variable. For example, suppose that a function f satisfies
fx (x, y) = cos(y)ex +2x+y 2 , fy (x, y) = − sin(y)ex +2xy+3, and f (0, 0) =
5.
a. Find all possible functions f of x and y such that fx (x, y) = cos(y)ex +
2x + y 2 . Your function will have both x and y as independent vari-
ables and may also contain summands that are functions of y alone.
b. Use the fact that fy (x, y) = − sin(y)ex + 2xy + 3 to determine any
unknown non-constant summands in your result from part (a).

c. Complete the problem by determining the specific function f that

satisfies the given conditions.

10.3 Second-Order Partial Derivatives

Motivating Questions
• Given a function f of two independent variables x and y, how are the
second-order partial derivatives of f defined?
• What do the second-order partial derivatives fxx , fyy , fxy , and fyx of a
function f tell us about the function’s behavior?
Recall that for a single-variable function f , the second derivative of f is
defined to be the derivative of the first derivative. That is, f 00 (x) = dx
d
[f 0 (x)],
which can be stated in terms of the limit definition of the derivative by writing

f 0 (x + h) − f 0 (x)
f 00 (x) = lim .
h→0 h
In what follows, we begin exploring the four different second-order partial
derivatives of a function of two variables and seek to understand what these
various derivatives tell us about the function’s behavior.
Preview Activity 10.3.1 Once again, let’s consider the function f defined
2
by f (x, y) = x sin(2y)
32 that measures a projectile’s range as a function of its
initial speed x and launch angle y. The graph of this function, including traces
with x = 150 and y = 0.6, is shown in Figure 10.3.1.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 125

z z
1500 1500

1000 1000

500 500

x x
200 200
y 150 y 150
1.5 100 1.5 100
1.0 50 1.0 50
0.5 0.5
0 0

Figure 10.3.1 The distance function with traces x = 150 and y = 0.6.

a. Compute the partial derivative fx . Notice that fx itself is a new function

of x and y, so we may now compute the partial derivatives of fx . Find
the partial derivative fxx = (fx )x and show that fxx (150, 0.6) ≈ 0.058.
b. Figure 10.3.2 shows the trace of f with y = 0.6 with three tangent lines
included. Explain how your result from part (b) of this preview activity
is reflected in this figure.
1000
f (x, 0.6)

800

600

400

200

x
50 100 150 200

Figure 10.3.2 The trace with y = 0.6.

c. Determine the partial derivative fy , and then find the partial derivative
fyy = (fy )y . Evaluate fyy (150, 0.6).
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 126

1000
f (150, y)

800

600

400

200

y
0.25 0.50 0.75 1.00 1.25

Figure 10.3.3 More traces of the range function.

d. Figure 10.3.3 shows the trace f (150, y) and includes three tangent lines.
Explain how the value of fyy (150, 0.6) is reflected in this figure.
e. Because fx and fy are each functions of both x and y, they each have
two partial derivatives. Not only can we compute fxx = (fx )x , but also
fxy = (fx )y ; likewise, in addition to fyy = (fy )y , but also fyx = (fy )x .
2
For the range function f (x, y) = x sin(2y)
32 , use your earlier computations
of fx and fy to now determine fxy and fyx . Write one sentence to explain
how you calculated these “mixed” partial derivatives.

10.3.1 Second-order Partial Derivatives

A function f of two independent variables x and y has two first order partial
derivatives, fx and fy . As we saw in Preview Activity 10.3.1, each of these
first-order partial derivatives has two partial derivatives, giving a total of four
second-order partial derivatives:
 
∂ ∂f ∂2f
• fxx = (fx )x = ∂x ∂x = ∂x2 ,
 
∂ ∂f ∂2f
• fyy = (fy )y = ∂y ∂y = ∂y 2 ,
 
∂ ∂f ∂2f
• fxy = (fx )y = ∂y ∂x = ∂y∂x ,
 
∂ ∂f ∂2f
• fyx = (fy )x = ∂x ∂y = ∂x∂y .

The first two are called unmixed second-order partial derivatives while the
last two are called the mixed second-order partial derivatives.
One aspect of this notation can be a little confusing. The notation

∂2f
 
∂ ∂f
=
∂y∂x ∂y ∂x

means that we first differentiate with respect to x and then with respect to y;
this can be expressed in the alternate notation fxy = (fx )y . However, to find
the second partial derivative
fyx = (fy )x
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 127

we first differentiate with respect to y and then x. This means that

∂2f ∂2f
= fxy , and = fyx .
∂y∂x ∂x∂y
Be sure to note carefully the difference between Leibniz notation and sub-
script notation and the order in which x and y appear in each. In addition,
remember that anytime we compute a partial derivative, we hold constant the
variable(s) other than the one we are differentiating with respect to.
Activity 10.3.2 Find all second order partial derivatives of the following func-
tions. For each partial derivative you calculate, state explicitly which variable
is being held constant.
a. f (x, y) = x2 y 3
b. f (x, y) = y cos(x)
c. g(s, t) = st3 + s4

d. How many second order partial derivatives does the function h defined
by h(x, y, z) = 9x9 z − xyz 9 + 9 have? Find hxz and hzx (you do not need
to find the other second order partial derivatives).
In Preview Activity 10.3.1 and Activity 10.3.2, you may have noticed that
the mixed second-order partial derivatives are equal. This observation holds
generally and is known as Clairaut’s Theorem.

Clairaut’s Theorem.
Let f be a function of several variables for which the partial derivatives
fxy and fyx are continuous near the point (a, b). Then

fxy (a, b) = fyx (a, b).

10.3.2 Interpreting the second-order Partial Derivatives

Recall from single variable calculus that the second derivative measures the
instantaneous rate of change of the derivative. This observation is the key to
understanding the meaning of the second-order partial derivatives.

z z z

6 6 6

4 4 4

2 2 2

-2 1 -2 1 -2 1
y -1 2 x y -1 2 x y -1 2 x
3 3 3

Figure 10.3.4 The tangent lines to a trace with increasing x.

Furthermore, we remember that the second derivative of a function at a
point provides us with information about the concavity of the function at that
point. Since the unmixed second-order partial derivative fxx requires us to
hold y constant and differentiate twice with respect to x, we may simply view
fxx as the second derivative of a trace of f where y is fixed. As such, fxx will
measure the concavity of this trace.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 128

Consider, for example, f (x, y) = sin(x)e−y . Figure 10.3.4 shows the graph
of this function along with the trace given by y = −1.5. Also shown are three
tangent lines to this trace, with increasing x-values from left to right among
the three plots in Figure 10.3.4.
That the slope of the tangent line is decreasing as x increases is reflected,
as it is in one-variable calculus, in the fact that the trace is concave down.
Indeed, we see that fx (x, y) = cos(x)e−y and so fxx (x, y) = − sin(x)e−y < 0,
since e−y > 0 for all values of y, including y = −1.5.
In the following activity, we further explore what second-order partial
derivatives tell us about the geometric behavior of a surface.
Activity 10.3.3 We continue to consider the function f defined by f (x, y) =
sin(x)e−y .
a. In Figure 10.3.5, we see the trace of f (x, y) = sin(x)e−y that has x held
constant with x = 1.75. We also see three different lines that are tangent
to the trace of f in the x direction at values of y that are increasing
from left to right in the figure. Write a couple of sentences that describe
whether the slope of the tangent lines to this curve increase or decrease as
y increases, and, after computing fyy (x, y), explain how this observation
is related to the value of fyy (1.75, y). Be sure to address the notion of
concavity in your response.(You need to be careful about the directions
in which x and y are increasing.)

z z z

6 6 6

4 4 4

2 2 2

-2 1 -2 1 -2 1
y -1 2 x y -1 2 x y -1 2 x
3 3 3

Figure 10.3.5 The tangent lines to a trace with increasing y.

b. In Figure 10.3.6, we start to think about the mixed partial derivative,

fxy . Here, we first hold y constant to generate the first-order partial
derivative fx , and then we hold x constant to compute fxy . This leads
to first thinking about a trace with x being constant, followed by slopes
of tangent lines in the y-direction that slide along the original trace. You
might think of sliding your pencil down the trace with x constant in a
way that its slope indicates (fx )y in order to further animate the three
snapshots shown in the figure.
z z z

6 6 6

4 4 4

2 2 2

-2 1 -2 1 -2 1
y -1 2 x y -1 2 x y -1 2 x
3 3 3

Figure 10.3.6 The trace of z = f (x, y) = sin(x)e−y with x = 1.75, along

with tangent lines in the y-direction at three different points.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 129

Based on Figure 10.3.6, is fxy (1.75, −1.5) positive or negative? Why?

c. Determine the formula for fxy (x, y), and hence evaluate fxy (1.75, −1.5).
How does this value compare with your observations in (b)?

d. We know that fxx (1.75, −1.5) measures the concavity of the y = −1.5
trace, and that fyy (1.75, −1.5) measures the concavity of the x = 1.75
trace. What do you think the quantity fxy (1.75, −1.5) measures?
e. On Figure 10.3.6, sketch the trace with y = −1.5, and sketch three tan-
gent lines whose slopes correspond to the value of fyx (x, −1.5) for three
different values of x, the middle of which is x = −1.5. Is fyx (1.75, −1.5)
positive or negative? Why? What does fyx (1.75, −1.5) measure?
Just as with the first-order partial derivatives, we can approximate second-
order partial derivatives in the situation where we have only partial information
about the function.
Activity 10.3.4 As we saw in Activity 10.2.5, the wind chill w(v, T ), in degrees
Fahrenheit, is a function of the wind speed, in miles per hour, and the air
temperature, in degrees Fahrenheit. Some values of the wind chill are recorded
in Table 10.3.7.
Table 10.3.7 Wind chill as a function of wind speed and temperature.
v\T -30 -25 -20 -15 -10 -5 0 5 10 15 20
5 -46 -40 -34 -28 -22 -16 -11 -5 1 7 13
10 -53 -47 -41 -35 -28 -22 -16 -10 -4 3 9
15 -58 -51 -45 -39 -32 -26 -19 -13 -7 0 6
20 -61 -55 -48 -42 -35 -29 -22 -15 -9 -2 4
25 -64 -58 -51 -44 -37 -31 -24 -17 -11 -4 3
30 -67 -60 -53 -46 -39 -33 -26 -19 -12 -5 1
35 -69 -62 -55 -48 -41 -34 -27 -21 -14 -7 0
40 -71 -64 -57 -50 -43 -36 -29 -22 -15 -8 -1

a. Estimate the partial derivatives wT (20, −15), wT (20, −10), and wT (20, −5).
Use these results to estimate the second-order partial wT T (20, −10).
b. In a similar way, estimate the second-order partial wvv (20, −10).
c. Estimate the partial derivatives wT (20, −10), wT (25, −10), and wT (15, −10),
and use your results to estimate the partial wT v (20, −10).

d. In a similar way, estimate the partial derivative wvT (20, −10).

e. Write several sentences that explain what the values wT T (20, −10), wvv (20, −10),
and wT v (20, −10) indicate regarding the behavior of w(v, T ).
As we have found in Activities 10.3.3 and Activity 10.3.4, we may think of
fxy as measuring the “twist” of the graph as we increase y along a particular
trace where x is held constant. In the same way, fyx measures how the graph
twists as we increase x. If we remember that Clairaut’s theorem tells us that
fxy = fyx , we see that the amount of twisting is the same in both directions.
This twisting is perhaps more easily seen in Figure 10.3.8, which shows the
graph of f (x, y) = −xy, for which fxy = −1.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 130

Figure 10.3.8 The graph of f (x, y) = −xy.

10.3.3 Summary
• There are four second-order partial derivatives of a function f of two
independent variables x and y:
fxx = (fx )x , fxy = (fx )y , fyx = (fy )x , and fyy = (fy )y .

• The unmixed second-order partial derivatives, fxx and fyy , tell us about
the concavity of the traces. The mixed second-order partial derivatives,
fxy and fyx , tell us how the graph of f twists.

10.3.4 Exercises
1. Calculate all four second-order partial derivatives of f (x, y) = 4x2 y+8xy 3 .
fxx (x, y) =
fxy (x, y) =
fyx (x, y) =
fyy (x, y) =
Answer 1. 8y
Answer 2. 4 · 2x + 8 · 3y 2
Answer 3. 4 · 2x + 8 · 3y 2
Answer 4. 8x · 3 · 2y
2. Find all the first and second order partial derivatives of f (x, y) = 3 sin(2x+
y) − 4 cos(x − y).
A. ∂f
∂x = fx =
∂f
B. ∂y = fy =
∂2f
C. ∂x2 = fxx =
∂2f
D. ∂y 2 = fy y =
∂2f
E. ∂x∂y = fy x =
∂2f
F. ∂y∂x = fxy =
Answer 1. 6 cos(2x + y) − (−4) sin(x − y)
Answer 2. 3 cos(2x + y) + (−4) sin(x − y)
Answer 3. −4 · 3 sin(2x + y) − (−4) cos(x − y)
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 131

Answer 4. −3 sin(2x + y) − (−4) cos(x − y)

Answer 5. −2 · 3 sin(2x + y) + (−4) cos(x − y)
Answer 6. −2 · 3 sin(2x + y) + (−4) cos(x − y)
3. Find the partial derivatives of the function

f (x, y) = xye4y

fx (x, y) =
fy (x, y) =
fxy (x, y) =
fyx (x, y) =
Answer 1. y exp(4y)
Answer 2. x(4y exp(4y) + exp(4y))
Answer 3. 4y exp(4y) + exp(4y)
Answer 4. 4y exp(4y) + exp(4y)
 
5x
4. Calculate all four second-order partial derivatives of f (x, y) = sin .
y
fxx (x, y) =
fxy (x, y) =
fyx (x, y) =
fyy (x, y) =
 
Answer 1. − 5y 5y
y 2 y 2 sin y
5x

2
   
Answer 2. 5y (y−5y·2y2) 2 cos 5x
y + 5y 5x
2
y y 2 sin 5x
y
 2    
5y 5y
Answer 3. − (y2 )2 cos 5x y − 5x
y 2 y 2 sin y
5x

    
Answer 4. − 5x 5x
y 2 y 2 sin y
5x
− 5x·2y
(y 2 )2
cos 5x y
4 5


5. Given F (r, s, t) = r 9s − t , compute:
Frst =
Answer. 0
6. Calculate all four second-order partial derivatives and check that fxy =
fyx . Assume the variables are restricted to a domain on which the function
is defined.
f (x, y) = e2xy
fxx =
fyy =
fxy =
fyx =
Answer 1. 4y 2 e2xy
Answer 2. 4x2 e2xy
Answer 3. 2e2xy + 4xye2xy
Answer 4. 2e2xy + 4xye2xy
7. Calculate all four second-order partial derivatives of f (x, y) = (2x + 4y) ey .
fxx (x, y) =
fxy (x, y) =
fyx (x, y) =
fyy (x, y) =
Answer 1. 0
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 132

Answer 2. 2ey
Answer 3. 2ey
Answer 4. 4ey + 4ey + (2x + 4y) ey
6
8. Let f (x, y) = (− (2x + y)) . Then
∂ 2f
∂x∂y =
∂ 3f
∂x∂y∂x =
∂ 3f
∂x2 ∂y =
4
Answer 1. 60(− (2x + y))
3
Answer 2. −480(− (2x + y))
3
Answer 3. −480(− (2x + y))
9. If zxy = 5y and all of the second order partial derivatives of z are contin-
uous, then
(a) zyx =
(b) zxyx =
(c) zxyy =
Answer 1. 5y
Answer 2. 0
Answer 3. 5
10. If z = f (x) + yg(x), what can we say about zyy ?
 zyy = 0

 zyy = y
 zyy = zxx
 zyy = g(x)

 We cannot say anything

11. Shown in Figure 10.3.9 is a contour plot of a function f with the values
of f labeled on the contours. The point (2, 1) is highlighted in red.

y
5
4
4

3
2
2

1
x
2 4

Figure 10.3.9 A contour plot of f (x, y).

a. Estimate the partial derivatives fx (2, 1) and fy (2, 1).

b. Determine whether the second-order partial derivative fxx (2, 1) is
positive or negative, and explain your thinking.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 133

c. Determine whether the second-order partial derivative fyy (2, 1) is

positive or negative, and explain your thinking.

d. Determine whether the second-order partial derivative fxy (2, 1) is

positive or negative, and explain your thinking.
e. Determine whether the second-order partial derivative fyx (2, 1) is
positive or negative, and explain your thinking.
f. Consider a function g of the variables x and y for which gx (2, 2) > 0
and gxx (2, 2) < 0. Sketch possible behavior of some contours around
(2, 2) on the left axes in Figure 10.3.10.

4 y 4 y

3 3

2 2

1 1

x x
1 2 3 4 1 2 3 4
Figure 10.3.10 Plots for contours of g and h.

g. Consider a function h of the variables x and y for which hx (2, 2) > 0

and hxy (2, 2) < 0. Sketch possible behavior of some contour lines
around (2, 2) on the right axes in Figure 10.3.10.
12. The Heat Index, I, (measured in apparent degrees F ) is a function of the
actual temperature T outside (in degrees F) and the relative humidity H
(measured as a percentage). A portion of the table which gives values for
this function, I(T, H), is reproduced in Table 10.3.11.
Table 10.3.11 Heat index.
T ↓\H → 70 75 80 85
90 106 109 112 115
92 112 115 119 123
94 118 122 127 132
96 125 130 135 141

a. State the limit definition of the value IT T (94, 75). Then, estimate
IT T (94, 75), and write one complete sentence that carefully explains
the meaning of this value, including units.

b. State the limit definition of the value IHH (94, 75). Then, estimate
IHH (94, 75), and write one complete sentence that carefully explains
the meaning of this value, including units.
c. Finally, do likewise to estimate IHT (94, 75), and write a sentence to
explain the meaning of the value you found.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 134

13. The temperature on a heated metal plate positioned in the first quadrant
of the xy-plane is given by
2
−(y−1)3
C(x, y) = 25e−(x−1) .

Assume that temperature is measured in degrees Celsius and that x

and y are each measured in inches.
a. Determine Cxx (x, y) and Cyy (x, y). Do not do any additional work
to algebraically simplify your results.
b. Calculate Cxx (1.1, 1.2). Suppose that an ant is walking past the
point (1.1, 1.2) along the line y = 1.2. Write a sentence to explain
the meaning of the value of Cxx (1.1, 1.2), including units.

c. Calculate Cyy (1.1, 1.2). Suppose instead that an ant is walking past
the point (1.1, 1.2) along the line x = 1.1. Write a sentence to
explain the meaning of the value of Cyy (1.1, 1.2), including units.
d. Determine Cxy (x, y) and hence compute Cxy (1.1, 1.2). What is the
meaning of this value? Explain, in terms of an ant walking on the
heated metal plate.
14. Let f (x, y) = 8 − x2 − y 2 and g(x, y) = 8 − x2 + 4xy − y 2 .
a. Determine fx , fy , fxx , fyy , fxy , and fyx .
b. Evaluate each of the partial derivatives in (a) at the point (0, 0).
c. What do the values in (b) suggest about the behavior of f near
(0, 0)? Plot a graph of f and compare what you see visually to what
the values suggest.
d. Determine gx , gy , gxx , gyy , gxy , and gyx .
e. Evaluate each of the partial derivatives in (d) at the point (0, 0).

f. What do the values in (e) suggest about the behavior of g near

(0, 0)? Plot a graph of g and compare what you see visually to what
the values suggest.
g. What do the functions f and g have in common at (0, 0)? What is
different? What do your observations tell you regarding the impor-
tance of a certain second-order partial derivative?
15. Let f (x, y) = 12 xy 2 represent the kinetic energy in Joules of an object of
mass x in kilograms with velocity y in meters per second. Let (a, b) be
the point (4, 5) in the domain of f .
2
a. Calculate ∂∂xf2 at the point (a, b). Then explain as best you can what
this second order partial derivative tells us about kinetic energy.
2
b. Calculate ∂∂yf2 at the point (a, b). Then explain as best you can what
this second order partial derivative tells us about kinetic energy.
2
∂ f
c. Calculate ∂y∂x at the point (a, b). Then explain as best you can what
this second order partial derivative tells us about kinetic energy.
2
∂ f
d. Calculate ∂x∂y at the point (a, b). Then explain as best you can what
this second order partial derivative tells us about kinetic energy.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 135

10.4 Linearization: Tangent Planes and Differ-

entials
Motivating Questions
• What does it mean for a function of two variables to be locally linear at
a point?
• How do we find the equation of the plane tangent to a locally linear
function at a point?

• What is the differential of a multivariable function of two variables and

what are its uses?
One of the central concepts in single variable calculus is that the graph of
a differentiable function, when viewed on a very small scale, looks like a line.
We call this line the tangent line and measure its slope with the derivative. In
this section, we will extend this concept to functions of several variables.
Let’s see what happens when we look at the graph of a two-variable function
on a small scale. To begin, let’s consider the function f defined by

x2
f (x, y) = 6 − − y2 ,
2
whose graph is shown in Figure 10.4.1.
6

4
z
2

-2

x
-2 y 2

Figure 10.4.1 The graph of f (x, y) = 6 − x2 /2 − y 2 .

We choose to study the behavior of this function near the point (x0 , y0 ) =
(1, 1). In particular, we wish to view the graph on an increasingly small scale
around this point, as shown in the two plots in Figure 10.4.2
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 136

6 6

4 4
z z
2 2

0 0.5
x x
0 y 2 0.5 y 1.5
Figure 10.4.2 The graph of f (x, y) = 6 − x2 /2 − y 2 .
Just as the graph of a differentiable single-variable function looks like a
line when viewed on a small scale, we see that the graph of this particular two-
variable function looks like a plane, as seen in Figure 10.4.3. In the following
preview activity, we explore how to find the equation of this plane.
6

4
z
2

0.5
x
0.5 y 1.5

Figure 10.4.3 The graph of f (x, y) = 6 − x2 /2 − y 2 .

In what follows, we will also use the important fact1 that the plane passing
through (x0 , y0 , z0 ) may be expressed in the form z = z0 + a(x − x0 ) + b(y − y0 ),
where a and b are constants.
x2
Preview Activity 10.4.1 Let f (x, y) = 6 − 2 − y 2 , and let (x0 , y0 ) = (1, 1).
2
a. Evaluate f (x, y) = 6 − x2 − y 2 and its partial derivatives at (x0 , y0 ); that
is, find f (1, 1), fx (1, 1), and fy (1, 1).
b. We know one point on the tangent plane; namely, the z-value of the
2
tangent plane agrees with the z-value on the graph of f (x, y) = 6− x2 −y 2
at the point (x0 , y0 ). In other words, both the tangent plane and the
1 As we saw in Section 9.5, the equation of a plane passing through the point (x , y , z )
0 0 0
may be written in the form A(x − x0 ) + B(y − y0 ) + C(z − z0 ) = 0. If the plane is not vertical,
then C 6= 0, and we can rearrange this and hence write C(z − z0 ) = −A(x − x0 ) − B(y − y0 )
and thus
A B
z = z0 − (x − x0 ) − (y − y0 )
C C
= z0 + a(x − x0 ) + b(y − y0 )
where a = −A/C and b = −B/C, respectively.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 137

graph of the function f contain the point (x0 , y0 , z0 ). Use this observation
to determine z0 in the expression z = z0 + a(x − x0 ) + b(y − y0 ).
x2
c. Sketch the traces of f (x, y) = 6 − 2 − y 2 for y = y0 = 1 and x = x0 = 1
below in Figure 10.4.4.
5.0 5.0
z = f (x, 1) z = f (1, y)

4.5 4.5

x y
4.0 4.0
0.5 1.0 1.5 0.5 1.0 1.5

Figure 10.4.4 The traces of f (x, y) with y = y0 = 1 and x = x0 = 1.

d. Determine the equation of the tangent line of the trace that you sketched
in the previous part with y = 1 (in the x direction) at the point x0 = 1.
6 6

4 4
z z
2 2

0.5 0.5
x x
0.5 y 1.5 0.5 y 1.5
Figure 10.4.5 The traces of f (x, y) and the tangent plane.

e. Figure 10.4.5 shows the traces of the function and the traces of the tan-
gent plane. Explain how the tangent line of the trace of f , whose equation
you found in the last part of this activity, is related to the tangent plane.
How does this observation help you determine the constant a in the equa-
tion for the tangent plane z = z0 + a(x − x0 ) + b(y − y0 )? (Hint: How do
you think fx (x0 , y0 ) should be related to zx (x0 , y0 )?)
f. In a similar way to what you did in (d), determine the equation of the
tangent line of the trace with x = 1 at the point y0 = 1. Explain how
this tangent line is related to the tangent plane, and use this observation
to determine the constant b in the equation for the tangent plane z =
z0 + a(x − x0 ) + b(y − y0 ). (Hint: How do you think fy (x0 , y0 ) should be
related to zy (x0 , y0 )?)
g. Finally, write the equation z = z0 + a(x − x0 ) + b(y − y0 ) of the tangent
plane to the graph of f (x, y) = 6−x2 /2−y 2 at the point (x0 , y0 ) = (1, 1).
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 138

10.4.1 The tangent plane

Before stating the formula for the equation of the tangent plane at a point for
a general function f = f (x, y), we need to discuss a technical condition. As we
have noted, when we look at the graph of a single-variable function on a small
scale near a point x0 , we expect to see a line; in this case, we say that f is
locally linear near x0 since the graph looks like a linear function locally around
x0 . Of course, there are functions, such as the absolute value function given
by f (x) = |x|, that are not locally linear at every point. In single-variable
calculus, we learn that if the derivative of a function exists at a point, then the
function is guaranteed to be locally linear there.
In a similar way, we say that a two-variable function f is locally linear near
(x0 , y0 ) provided that the graph of f looks like a plane (its tangent plane) when
viewed on a small scale near (x0 , y0 ). How can we tell when a function of two
variables is locally linear at a point?
It is not unreasonable to expect that if fx (a, b) and fy (a, b) exist for some
function f at a point (a, b), then f is locally linear at (a, b). This is not
sufficient, however. As an example, consider the function f defined by f (x, y) =
x1/3 y 1/3 . In Exercise 10.4.5.11 you are asked to show that fx (0, 0) and fy (0, 0)
both exist, but that f is not locally linear at (0, 0) (see Figure 10.4.12). So the
existence of the two first order partial derivatives at a point does not guarantee
local linearity at that point.
It would take us too far afield to provide a rigorous dicussion of differentia-
bility of functions of more than one variable (see Exercise 10.4.5.15) for a little
more detail, so we will be content to define stronger, but more easily verified,
conditions that ensure local linearity.

Differentiablity.

If f is a function of the independent variables x and y and both fx

and fy exist and are continuous in an open disk containing the point
(x0 , y0 ), then f is continuously differentiable at (x0 , y0 ).

As a consequence, whenever a function z = f (x, y) is continuously differ-

entiable at a point (x0 , y0 ), it follows that the function has a tangent plane at
(x0 , y0 ). Viewed up close, the tangent plane and the function are then virtually
indistinguishable. (We won’t formally define differentiability of multivariable
functions here, and for our purposes continuous differentiability is the only
condition we will ever need to use. It is important to note that continuous
differentiability is a stronger condition than differentiability. All of the results
we encounter will apply to differentiable functions, and so also apply to con-
tinuously differentiable functions.) In addition, as in Preview Activity 10.4.1,
we find the following general formula for the tangent plane.

The tangent plane.

If f (x, y) has continuous first-order partial derivatives, then the equa-

tion of the plane tangent to the graph of f at the point (x0 , y0 , f (x0 , y0 ))
is

z = f (x0 , y0 ) + fx (x0 , y0 )(x − x0 ) + fy (x0 , y0 )(y − y0 ). (10.4.1)

Important Note: As can be seen in Exercise 10.4.5.11, it is possible that

fx (x0 , y0 ) and fy (x0 , y0 ) can exist for a function f , and so the plane z =
f (x0 , y0 ) + fx (x0 , y0 )(x − x0 ) + fy (x0 , y0 )(y − y0 ) exists even though f is not
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 139

locally linear at (x0 , y0 ) (because the graph of f does not look linear when we
zoom in around the point (x0 , y0 )). In such a case this plane is not tangent to
the graph. Differentiability for a function of two variables implies the existence
of a tangent plane, but the existence of the two first order partial derivatives
of a function at a point does not imply differentiaility. This is quite different
than what happens in single variable calculus.
Finally, one important note about the form of the equation for the tangent
plane, z = f (x0 , y0 ) + fx (x0 , y0 )(x − x0 ) + fy (x0 , y0 )(y − y0 ). Say, for exam-
ple, that we have the particular tangent plane z = 7 − 2(x − 3) + 4(y + 1).
Observe that we can immediately read from this form that fx (3, −1) = −2
and fy (3, −1) = 4; furthermore, fx (3, −1) = −2 is the slope of the trace to
both f and the tangent plane in the x-direction at (−3, 1). In the same way,
fy (3, −1) = 4 is the slope of the trace of both f and the tangent plane in the
y-direction at (3, −1).
Activity 10.4.2
a. Find the equation of the tangent plane to f (x, y) = 2 + 4x − 3y at the
point (1, 2). Simplify as much as possible. Does the result surprise you?
Explain.

b. Find the equation of the tangent plane to f (x, y) = x2 y at the point

(1, 2).

10.4.2 Linearization
In single variable calculus, an important use of the tangent line is to approx-
imate the value of a differentiable function. Near the point x0 , the tangent
line to the graph of f at x0 is close to the graph of f near x0 , as shown in
Figure 10.4.6.

y y

y = L(x) y = L(x)

y = f (x) y = f (x)

x x
x0 x0

Figure 10.4.6 The linearization of the single-variable function f (x).

In this single-variable setting, we let L denote the function whose graph is
the tangent line, and thus

L(x) = f (x0 ) + f 0 (x0 )(x − x0 )

Furthermore, observe that f (x) ≈ L(x) near x0 . We call L the linearization

of f .
In the same way, the tangent plane to the graph of a differentiable function
z = f (x, y) at a point (x0 , y0 ) provides a good approximation of f (x, y) near
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 140

(x0 , y0 ). Here, we define the linearization, L, to be the two-variable function

whose graph is the tangent plane, and thus
L(x, y) = f (x0 , y0 ) + fx (x0 , y0 )(x − x0 ) + fy (x0 , y0 )(y − y0 ).
Finally, note that f (x, y) ≈ L(x, y) for points near (x0 , y0 ). This is illus-
trated in Figure 10.4.7.
6
z = L(x, y)

4
z
2

z = f (x, y)

0.5
x
0.5 y 1.5

Figure 10.4.7 The linearization of f (x, y).

Activity 10.4.3 In what follows, we find the linearization of several different
functions that are given in algebraic, tabular, or graphical form.
a. Find the linearization L(x, y) for the function g defined by
x
g(x, y) =
x2 + y2
at the point (1, 2). Then use the linearization to estimate the value of
g(0.8, 2.3).
b. Table 10.4.8 provides a collection of values of the wind chill w(v, T ), in
degrees Fahrenheit, as a function of wind speed, in miles per hour, and
temperature, also in degrees Fahrenheit.
Table 10.4.8 Wind chill as a function of wind speed and tem-
perature.
v\T −30 −25 −20 −15 −10 −5 0 5 10 15 20
5 −46 −40 −34 −28 −22 −16 −11 −5 1 7 13
10 −53 −47 −41 −35 −28 −22 −16 −10 −4 3 9
15 −58 −51 −45 −39 −32 −26 −19 −13 −7 0 6
20 −61 −55 −48 −42 −35 −29 −22 −15 −9 −2 4
25 −64 −58 −51 −44 −37 −31 −24 −17 −11 −4 3
30 −67 −60 −53 −46 −39 −33 −26 −19 −12 −5 1
35 −69 −62 −55 −48 −41 −34 −27 −21 −14 −7 0
40 −71 −64 −57 −50 −43 −36 −29 −22 −15 −8 −1

Use the data to first estimate the appropriate partial derivatives, and
then find the linearization L(v, T ) at the point (20, −10). Finally, use
the linearization to estimate w(10, −10), w(20, −12), and w(18, −12).
Compare your results to what you obtained in Activity 10.2.5
c. Figure 10.4.9 gives a contour plot of a continuously differentiable function
f.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 141

y
5
4
4

3
2
2

1
x
2 4

Figure 10.4.9 A contour plot of f (x, y).

After estimating appropriate partial derivatives, determine the lineariza-

tion L(x, y) at the point (2, 1), and use it to estimate f (2.2, 1), f (2, 0.8),
and f (2.2, 0.8).

10.4.3 Differentials
As we have seen, the linearization L(x, y) enables us to estimate the value of
f (x, y) for points (x, y) near the base point (x0 , y0 ). Sometimes, however, we
are more interested in the change in f as we move from the base point (x0 , y0 )
to another point (x, y).
z

df
∆f

y
(x0 , y0 )

dx = ∆x

dy = ∆y (x, y)
x

Figure 10.4.10 The differential df approximates the change in f (x, y).

Figure 10.4.10 illustrates this situation. Suppose we are at the point (x0 , y0 ),
and we know the value f (x0 , y0 ) of f at (x0 , y0 ). If we consider the displace-
ment h∆x, ∆yi to a new point (x, y) = (x0 + ∆x, y0 + ∆y), we would like to
know how much the function has changed. We denote this change by ∆f ,
where
∆f = f (x, y) − f (x0 , y0 ).
A simple way to estimate the change ∆f is to approximate it by df , which
represents the change in the linearization L(x, y) as we move from (x0 , y0 ) to
(x, y). This gives
∆f ≈ df = L(x, y) − f (x0 , y0 )
= [f (x0 , y0 ) + fx (x0 , y0 )(x − x0 ) + fy (x0 , y0 )(y − y0 )] − f (x0 , y0 )
= fx (x0 , y0 )∆x + fy (x0 , y0 )∆y.
For consistency, we will denote the change in the independent variables as
dx = ∆x and dy = ∆y, and thus
∆f ≈ df = fx (x0 , y0 ) dx + fy (x0 , y0 ) dy. (10.4.2)
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 142

Expressed equivalently in Leibniz notation, we have

∂f ∂f
df = dx + dy. (10.4.3)
∂x ∂y
We call the quantities dx, dy, and df differentials, and we think of them
as measuring small changes in the quantities x, y, and f . Equations (10.4.2)
and (10.4.3) express the relationship between these changes. Equation (10.4.3)
resembles an important idea from single-variable calculus: when y depends on
x, it follows in the notation of differentials that
dy
dy = y 0 dx = dx.
dx
We will illustrate the use of differentials with an example.
Example 10.4.11 Suppose we have a machine that manufactures rectangles
of width x = 20 cm and height y = 10 cm. However, the machine isn’t perfect,
and therefore the width could be off by dx = ∆x = 0.2 cm and the height
could be off by dy = ∆y = 0.4 cm.
The area of the rectangle is

A(x, y) = xy,

so that the area of a perfectly manufactured rectangle is A(20, 10) = 200 square
centimeters. Since the machine isn’t perfect, we would like to know how much
the area of a given manufactured rectangle could differ from the perfect rect-
angle. We will estimate the uncertainty in the area using (10.4.2), and find
that
∆A ≈ dA = Ax (20, 10) dx + Ay (20, 10) dy.
Since Ax = y and Ay = x, we have

∆A ≈ dA = 10 dx + 20 dy = 10 · 0.2 + 20 · 0.4 = 10.

That is, we estimate that the area in our rectangles could be off by as much
as 10 square centimeters. 
Activity 10.4.4 The questions in this activity explore the differential in several
different contexts.
a. Suppose that the elevation of a landscape is given by the function h, where
we additionally know that h(3, 1) = 4.35, hx (3, 1) = 0.27, and hy (3, 1) =
−0.19. Assume that x and y are measured in miles in the easterly and
northerly directions, respectively, from some base point (0, 0). Your GPS
device says that you are currently at the point (3, 1). However, you
know that the coordinates are only accurate to within 0.2 units; that is,
dx = ∆x = 0.2 and dy = ∆y = 0.2. Estimate the uncertainty in your
elevation using differentials.
b. The pressure, volume, and temperature of an ideal gas are related by the
equation
P = P (T, V ) = 8.31T /V,
where P is measured in kilopascals, V in liters, and T in kelvin. Find
the pressure when the volume is 12 liters and the temperature is 310 K.
Use differentials to estimate the change in the pressure when the volume
increases to 12.3 liters and the temperature decreases to 305 K.
c. Refer to Table 10.4.8, the table of values of the wind chill w(v, T ), in
degrees Fahrenheit, as a function of temperature, also in degrees Fahren-
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 143

heit, and wind speed, in miles per hour. Suppose your anemometer says
the wind is blowing at 25 miles per hour and your thermometer shows a
reading of −15◦ degrees. However, you know your thermometer is only
accurate to within 2◦ degrees and your anemometer is only accurate to
within 3 miles per hour. What is the wind chill based on your mea-
surements? Estimate the uncertainty in your measurement of the wind
chill.

10.4.4 Summary
• A function f of two independent variables is locally linear at a point
(x0 , y0 ) if the graph of f looks like a plane as we zoom in on the graph
around the point (x0 , y0 ). In this case, the equation of the tangent plane
is given by

z = f (x0 , y0 ) + fx (x0 , y0 )(x − x0 ) + fy (x0 , y0 )(y − y0 ).

• The tangent plane L(x, y) = f (x0 , y0 )+fx (x0 , y0 )(x−x0 )+fy (x0 , y0 )(y −
y0 ), when considered as a function, is called the linearization of a differ-
entiable function f at (x0 , y0 ) and may be used to estimate values of
f (x, y); that is, f (x, y) ≈ L(x, y) for points (x, y) near (x0 , y0 ).
• A function f of two independent variables is differentiable at (x0 , y0 )
provided that both fx and fy exist and are continuous in an open disk
containing the point (x0 , y0 ).
• The differential df of a function f = f (x, y) is related to the differentials
dx and dy by
df = fx (x0 , y0 )dx + fy (x0 , y0 )dy.
We can use this relationship to approximate small changes in f that
result from small changes in x and y.

10.4.5 Exercises
p
1. Find the linearization L (x, y) of the function f (x, y) = 161 − 9x2 − 4y 2
at (−4, −2).
L (x, y) =
Note: Your answer should be an expression in x and y; e.g. “3x - 5y
+ 9”
Answer. 36(x − (−4)) + 8(y − (−2)) + 1
2. Find the equation of the tangent plane to the surface z = e2x/17 ln (1y) at
the point (3, 2, 0.9865).
z=
Note: Your answer should be an expression of x and y; e.g. “5x + 2y
- 3”
Answer. 0.116061x + 0.711624y + (−0.784911)
3. A student was asked to find the equation of the tangent plane to the
surface z = x4 − y 5 at the point
(x, y) = (4, 5). The student’s answer was
z = −2869 + 4x3 (x − 4) − 5y 4 (y − 5).
(a) At a glance, how do you know this is wrong. What mistakes did
the student make? Select all that apply.
 The answer is not a linear function.  The (x - 4) and (y -
5) should be x and y.  The partial derivatives were not evaluated
a the point.  The -2869 should not be in the answer.  All
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 144

of the above
(b) Find the correct equation for the tangent plane.
z=
Answer. AC
4. (a) Check the local linearity of f (x, y) = e−x cos(y) near x = 1, y = 1.5
by filling in the following table of values of f for x = 0.9, 1, 1.1 and
y = 1.4, 1.5, 1.6. Express values of f with 4 digits after the decimal
point.
x= 0.9 1 1.1
y = 1.4
y = 1.5
y = 1.6
(b) Next, fill in the table for the values x = 0.99, 1, 1.01 and y =
1.49, 1.5, 1.51, again showing 4 digits after the decimal point.

x= 0.99 1 1.01
y = 1.49
y = 1.5
y = 1.51
Notice if the two tables look nearly linear, and whether the second
looks more linear than the first (in particular, think about how you would
decide if they were linear, or if the one were more closely linear than the
other).
(c) Give the local linearization of f (x, y) = e−x cos(y) at (1, 1.5):
Using the second of your tables:
f (x, y) ≈
Using the fact that fx (x, y) = −e−x cos(y) and fy (x, y) = −e−x sin(y):
f (x, y) ≈
Answer 1. 0.0691
Answer 2. 0.0625
Answer 3. 0.0566
Answer 4. 0.0288
Answer 5. 0.026
Answer 6. 0.0235
Answer 7. −0.0119
Answer 8. −0.0107
Answer 9. −0.0097
Answer 10. 0.03
Answer 11. 0.0297
Answer 12. 0.0294
Answer 13. 0.0263
Answer 14. 0.026
Answer 15. 0.0258
Answer 16. 0.0226
Answer 17. 0.0224
Answer 18. 0.0221
Answer 19. 0.026 + (0.0258−0.0263)(x−1)
0.02 + (0.0224−0.0297)(y−1.5)
0.02
Answer 20. 0.0260228 + −0.0260228(x − 1) + −0.366958(y − 1.5)
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 145

5. Suppose that z is a linear function of x and y with slope -5 in the x

direction and slope 5 in the y direction.
(a) A change of 0.2 in x and 0.5 in y produces what change in z?
change in z =
(b) If z = 6 when x = 3 and y = 2, what is the value of z when x = 2.7
and y = 1.9?
z=
Answer 1. −5 · 0.2 + 5 · 0.5
Answer 2. 6 + −5 · (2.7 − 3) + 5 · (1.9 − 2)
6. Find the differential of the function w = x3 sin(y 5 z 1 )
dw = dx+ dy+
dz


Answer 1. 3x2 sin y 5 z 1


Answer 2. 5x3 y 4 z 1 cos y 5 z 1


Answer 3. 1x3 y 5 z 0 cos y 5 z 1
7. The dimensions of a closed rectangular box are measured as 60 centime-
ters, 60 centimeters, and 80 centimeters, respectively, with the error in
each measurement at most .2 centimeters. Use differentials to estimate
the maximum error in calculating the surface area of the box.
square centimeters
Answer. 160
8. One mole of ammonia gas is contained in a vessel which is capable of
changing its volume (a compartment sealed by a piston, for example).
The total energy U (in Joules) of the ammonia is a function of the volume
V (in cubic meters) of the container, and the temperature T (in degrees
Kelvin) of the gas. The differential dU is given by dU = 840dV +27.32dT .
(a) How does the energy change if the volume is held constant and the
temperature is decreased slightly?

 it increases slightly
 it does not change
 it decreases slightly

(b) How does the energy change if the temperature is held constant
and the volume is increased slightly?

 it does not change

 it increases slightly

 it decreases slightly

(c) Find the approximate change in energy if the gas is compressed by

150 cubic centimeters and heated by 3 degrees Kelvin.
Change in energy = .
Please include units in your answer.
Answer. C
9. An unevenly heated metal plate has temperature T (x, y) in degrees Celsius
at a point (x, y). If T (2, 1) = 119, Tx (2, 1) = 19, and Ty (2, 1) = −14,
estimate the temperature at the point (2.04, 0.96).
T (2.04, 0.96) ≈ .
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 146

Please include units in your answer.

Answer. 120.32 degC
10. Let f be the function defined by f (x, y) = 2x2 + 3y 3 .
a. Find the equation of the tangent plane to f at the point (1, 2).

b. Use the linearization to approximate the values of f at the points

(1.1, 2.05) and (1.3, 2.2).
c. Compare the approximations form part (b) to the exact values of
f (1.1, 2.05) and f (1.3, 2.2). Which approximation is more accurate.
Explain why this should be expected.
11. Let f be the function defined by f (x, y) = x1/3 y 1/3 , whose graph is shown
in Figure 10.4.12.

0
-1

x
0
-1
-1 0 y 1

Figure 10.4.12 The surface for f (x, y) = x1/3 y 1/3 .

a. Determine
f (0 + h, 0) − f (0, 0)
lim .
h→0 h
What does this limit tell us about fx (0, 0)?

b. Note that f (x, y) = f (y, x), and this symmetry implies that fx (0, 0) =
fy (0, 0). So both partial derivatives of f exist at (0, 0). A picture of
the surface defined by f near (0, 0) is shown in Figure 10.4.12. Based
on this picture, do you think f is locally linear at (0, 0)? Why?

c. Show that the curve where x = y on the surface defined by f is not

differentiable at 0. What does this tell us about the local linearity
of f at (0, 0)?
x2
d. Is the function f defined by f (x, y) = y 2 +1 locally linear at (0, 0)?
Why or why not?
12. Let g be a function that is differentiable at (−2, 5) and suppose that its
tangent plane at this point is given by z = −7 + 4(x + 2) − 3(y − 5).
a. Determine the values of g(−2, 5), gx (−2, 5), and gy (−2, 5). Write
one sentence to explain your thinking.
b. Estimate the value of g(−1.8, 4.7). Clearly show your work and
thinking.

c. Given changes of dx = −0.34 and dy = 0.21, estimate the corre-

CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 147

sponding change in g that is given by its differential, dg.

d. Suppose that another function h is also differentiable at (−2, 5),
but that its tangent plane at (−2, 5) is given by 3x + 2y − 4z = 9.
Determine the values of h(−2, 5), hx (−2, 5), and hy (−2, 5), and then
estimate the value of h(−1.8, 4.7). Clearly show your work and
thinking.
13. In the following questions, we determine and apply the linearization for
several different functions.
a. Find the linearization L(x, y) for the function f defined by f (x, y) =
cos(x)(2e2y + e−2y ) at the point (x0 , y0 ) = (0, 0). Use the lineariza-
tion to estimate the value of f (0.1, 0.2). Compare your estimate to
the actual value of f (0.1, 0.2).
b. The Heat Index, I, (measured in apparent degrees F) is a function
of the actual temperature T outside (in degrees F) and the relative
humidity H (measured as a percentage). A portion of the table
which gives values for this function, I = I(T, H), is provided in
Table 10.4.13.
Table 10.4.13 Heat index.
T ↓\H → 70 75 80 85
90 106 109 112 115
92 112 115 119 123
94 118 122 127 132
96 125 130 135 141

Suppose you are given that IT (94, 75) = 3.75 and IH (94, 75) = 0.9.
Use this given information and one other value from the table to es-
timate the value of I(93.1, 77) using the linearization at (94, 75). Us-
ing proper terminology and notation, explain your work and think-
ing.
c. Just as we can find a local linearization for a differentiable function
of two variables, we can do so for functions of three or more variables.
By extending the concept of the local linearization from two to three
variables, find the linearization of the function h(x, y, z) = e2x (y +
z 2 ) at the point (x0 , y0 , z0 ) = (0, 1, −2). Then, use the linearization
to estimate the value of h(−0.1, 0.9, −1.8).
14. In the following questions, we investigate two different applied settings
using the differential.
a. Let f represent the vertical displacement in centimeters from the rest
position of a string (like a guitar string) as a function of the distance
x in centimeters from the fixed left end of the string and y the time
in seconds after the string has been plucked. (An interesting video of
this can be seen at https://www.youtube.com/watch?v=TKF6nFzpHBUA.)
A simple model for f could be

f (x, y) = cos(x) sin(2y).

Use the differential to approximate how much more this vibrating

string is vertically displaced from its position at (a, b) = π4 , π3 if

we decrease a by 0.01 cm and increase the time by 0.1
seconds.

Compare to the value of f at the point π4 − 0.01, π3 + 0.1 .
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 148

b. Resistors used in electrical circuits have colored bands painted on

them to indicate the amount of resistance and the possible error in
the resistance. When three resistors, whose resistances are R1 , R2 ,
and R3 , are connected in parallel, the total resistance R is given by
1 1 1 1
= + + .
R R1 R2 R3
Suppose that the resistances are R1 = 25Ω, R2 = 40Ω, and R3 =
50Ω. Find the total resistance R. If you know each of R1 , R2 , and
R3 with a possible error of 0.5%, estimate the maximum error in
your calculation of R.
15. In this section we argued that if f = f (x, y) is a function of two variables
and if fx and fy both exist and are continuous in an open disk contain-
ing the point (x0 , y0 ), then f is differentiable at (x0 , y0 ). This condition
ensures that f is differentiable at (x0 , y0 ), but it does not define what it
means for f to be differentiable at (x0 , y0 ). In this exercise we explore the
definition of differentiability of a function of two variablesp in more detail.
Throughout, let g be the function defined by g(x, y) = |xy|.
a. Use appropriate technology to plot the graph of g on the domain
[−1, 1] × [−1, 1]. Explain why g is not locally linear at (0, 0).
b. Show that both gx (0, 0) and gy (0, 0) exist. If g is locally linear at
(0, 0), what must be the equation of the tangent plane L to g at
(0, 0)?
c. Recall that if a function f = f (x) of a single variable is differentiable
at x = x0 , then

f (x0 + h) − f (x0 )
f 0 (x0 ) = lim
h→0 h
exists. We saw in single variable calculus that the existence of f 0 (x0 )
means that the graph of f is locally linear at x = x0 . In other
words, the graph of f looks like its linearization L(x) = f (x0 ) +
f 0 (x0 )(x − x0 ) for x close to x0 . That is, the values of f (x) can be
closely approximated by L(x) as long as x is close to x0 . We can
measure how good the approximation of L(x) is to f (x) with the
error function

E(x) = L(x) − f (x) = f (x0 ) + f 0 (x0 )(x − x0 ) − f (x).

As x approaches x0 , E(x) approaches f (x0 ) + f 0 (x0 )(0) − f (x0 ) = 0,

and so L(x) provides increasingly better approximations p to f (x) as
x gets closer to x0 . Show that, even though g(x, y) = |xy| is not
locally linear at (0, 0), its error term

E(x, y) = L(x, y) − g(x, y)

at (0, 0) has a limit of 0 as (x, y) approaches (0, 0). (Use the lin-
earization you found in part (b).) This shows that just because an
error term goes to 0 as (x, y) approaches (x0 , y0 ), we cannot conclude
that a function is locally linear at (x0 , y0 ).
d. As the previous part illustrates, having the error term go to 0 does
not ensure that a function of two variables is locally linear. Instead,
we need a notation of a relative error. To see how this works, let
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 149

us return to the single variable case for a moment and consider

f = f (x) as a function of one variable. If we let x = x0 + h,
where |h| is the distance from x to x0 , then the relative error in
approximating f (x0 + h) with L(x0 + h) is

E(x0 + h)
.
h
Show that, for a function f = f (x) of a single variable, the limit of
the relative error is 0 as h approaches 0.
e. Even though the error term for a function of two variables might
have a limit of 0 at a point, our example shows that the function
may not be locally linear at that point. So we use the concept of
relative error to define differentiability of a function of two variables.
When we consider differentiability of a function f = f (x, y) at a
point (x0 , y0 ), then √if x = x0 + h and y = y0 + k, the distance from
(x, y) to (x0 , y0 ) is h2 + k 2 .
Definition 10.4.14 A function f = f (x, y) is differentiable at a
point (x0 , y0 ) if there is a linear function L = L(x, y) = f (x0 , y0 ) +
m(x − x0 ) + n(y − y0 ) such that the relative error

E(x0 + h, y0 + k)
√ ,
h2 + k 2
has at limit of 0 at (h, k) = (0, 0), where E(x, y) = f (x, y) − L(x, y),
h = x − x0 , and k = y − y0 . ♦
A function f is differentiable if it is differentiable at every point in
its domain. The function L in thepdefinition is the linearization of f
at (x0 , y0 ). Verify that g(x, y) = |xy| is not differentiable at (0, 0)
by showing that the relative error at (0, 0) does not have a limit at
(0, 0). Conclude that the existence of partial derivatives at a point is
not enough to ensure differentiability at that point. (Hint: Consider
the limit along different paths.)
16. Suppose that a function f = f (x, y) is differentiable at a point (x0 , y0 ).
Let L = L(x, y) = f (x0 , y0 ) + m(x − x0 ) + n(y − y0 ) as in the conditions of
Definition 10.4.14. Show that m = fx (x0 , y0 ) and n = fy (x0 , y0 ). (Hint:
Calculate the limits of the relative errors when h = 0 and k = 0.)
17. We know that if a function of a single variable is differentiable at a point,
then that function is also continuous at that point. In this exercise we
determine that the same property holds for functions of two variables. A
function f of the two variables x and y is continuous at a point (x0 , y0 )
in its domain if
lim f (x, y) = f (x0 , y0 )
(x,y)→(x0 ,y0 )

or (letting x = x0 + h and y = y0 + k,

lim f (x0 + h, y + k) = f (x0 , y0 ).

(h,k)→(0,0)

Show that if f is differentiable at (x0 , y0 ), then f is continuous at (x0 , y0 ).

(Hint: Multiply both
√ sides of the equality that comes from differentiability
by lim(h,k)→(0,0) h2 + k 2 .)
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 150

10.5 The Chain Rule

Motivating Questions
• What is the Chain Rule and how do we use it to find a derivative?
• How can we use a tree diagram to guide us in applying the Chain Rule?
In single-variable calculus, we encountered situations in which some quan-
tity z depends on y and, in turn, y depends on x. A change in x produces a
change in y, which consequently produces a change in z. Using the language
of differentials that we saw in the previous section, these changes are naturally
related by
dz dy
dz = dy and dy = dx.
dy dx
In terms of instantaneous rates of change, we then have
dz dy dz
dz = dx = dx
dy dx dx
and thus
dz dz dy
= .
dx dy dx
This most recent equation we call the Chain Rule.
In the case of a function f of two variables where z = f (x, y), it might be
that both x and y depend on another variable t. A change in t then produces
changes in both x and y, which then cause z to change. In this section we
will see how to find the change in z that is caused by a change in t, leading us
to multivariable versions of the Chain Rule involving both regular and partial
derivatives.
Preview Activity 10.5.1 Suppose you are driving around in the xy-plane in
such a way that your position r(t) at time t is given by function

r(t) = hx(t), y(t)i = h2 − t2 , t3 + 1i.

The path taken is shown on the left of Figure 10.5.1.

3 y z

1
y

x
x
1 2 3

Figure 10.5.1 Left: Your position in the plane. Right: The corresponding
temperature.
Suppose, furthermore, that the temperature at a point in the plane is given
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 151

by
1 1
T (x, y) = 10 − x2 − y 2 ,
2 5
and note that the surface generated by T is shown on the right of Figure 10.5.1.
Therefore, as time passes, your position (x(t), y(t)) changes, and, as your po-
sition changes, the temperature T (x, y) also changes.
a. The position function r provides a parameterization x = x(t) and y = y(t)
of the position at time t. By substituting x(t) for x and y(t) for y in the
formula for T , we can write T = T (x(t), y(t)) as a function of t. Make
these substitutions to write T as a function of t and then use the Chain
Rule from single variable calculus to find dT
dt . (Do not do any algebra to
simplify the derivative, either before taking the derivative, nor after.)

b. Now we want to understand how the result from part (a) can be obtained
from T as a multivariable function. Recall from the previous section that
small changes in x and y produce a change in T that is approximated by

∆T ≈ Tx ∆x + Ty ∆y.

The Chain Rule tells us about the instantaneous rate of change of T , and
this can be found as
∆T Tx ∆x + Ty ∆y
lim = lim . (10.5.1)
∆t→0 ∆t ∆t→0 ∆t
Use equation (10.5.1) to explain why the instantaneous rate of change of
T that results from a change in t is
dT ∂T dx ∂T dy
= + . (10.5.2)
dt ∂x dt ∂y dt

c. Using the original formulas for T , x, and y in the problem statement,

calculate all of the derivatives in Equation (10.5.2) (with Tx and Ty in
terms of x and y, and x0 and y 0 in terms of t), and hence write the
right-hand side of Equation (10.5.2) in terms of x, y, and t.
d. Compare the results of parts (a) and (c). Write a couple of sentences that
identify specifically how each term in (c) relates to a corresponding terms
in (a). This connection between parts (a) and (c) provides a multivariable
version of the Chain Rule.

10.5.1 The Chain Rule

As Preview Activity 10.3.1 suggests, the following version of the Chain Rule
holds in general.

The Chain Rule.

Let z = f (x, y), where f is a differentiable function of the independent
variables x and y, and let x and y each be differentiable functions of an
independent variable t. Then
dz ∂z dx ∂z dy
= + . (10.5.3)
dt ∂x dt ∂y dt

It is important to note the differences among the derivatives in (10.5.3).

CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 152

Since z is a function of the two variables x and y, the derivatives in the Chain
Rule for z with respect to x and y are partial derivatives. However, since
x = x(t) and y = y(t) are functions of the single variable t, their derivatives
are the standard derivatives of functions of one variable. When we compose z
with x(t) and y(t), we then have z as a function of the single variable t, making
the derivative of z with respect to t a standard derivative from single variable
calculus as well.
To understand why this Chain Rule works in general, suppose that some
quantity z depends on x and y so that
∂z ∂z
dz = dx + dy. (10.5.4)
∂x ∂y
Next, suppose that x and y each depend on another quantity t, so that
dx dy
dx = dt and dy = dt. (10.5.5)
dt dt
Combining Equations (10.5.4) and (10.5.5), we find that
∂z dx ∂z dy dz
dz = dt + dt = dt,
∂x dt ∂y dt dt
which is the Chain Rule in this particular context, as expressed in Equa-
tion (10.5.3).
Activity 10.5.2 In the following questions, we apply the Chain Rule in several
different contexts.
a. Suppose that we have a function z defined by z(x, y) = x2 + xy 3 . In
addition, suppose that x and y are restricted to points that move around
the plane by following a circle of radius 2 centered at the origin that is
parameterized by

x(t) = 2 cos(t), and y(t) = 2 sin(t).

i. Use the Chain Rule to find the resulting instantaneous rate of change
dz
dt .
ii. Substitute x(t) for x and y(t) for y in the rule for z to write z in
terms of t and calculate dz
dt directly. Compare to the result of part
(i.).

b. Suppose that the temperature on a metal plate is given by the function

T with
T (x, y) = 100 − (x2 + 4y 2 ),
where the temperature is measured in degrees Fahrenheit and x and y
are each measured in feet.
i. Find Tx and Ty . What are the units on these partial derivatives?
ii. Suppose an ant is walking along the x-axis at the rate of 2 feet per
minute toward the origin. When the ant is at the point (2, 0), what
is the instantaneous rate of change in the temperature dT /dt that
the ant experiences. Include units on your response.
iii. Suppose instead that the ant walks along an ellipse with x = 6 cos(t)
and y = 3 sin(t), where t is measured in minutes. Find dT dt at t =
π/6, t = π/4, and t = π/3. What does this seem to tell you about
the path along which the ant is walking?
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 153

c. Suppose that you are walking along a surface whose elevation is given
by a function f . Furthermore, suppose that if you consider how your
location corresponds to points in the xy-plane, you know that when you
pass the point (2, 1), your velocity vector is v = h−1, 2i. If some contours
of f are as shown in Figure 10.5.2, estimate the rate of change df /dt when
you pass through (2, 1).

y
5
4
4

3
2
2

1
x
2 4

Figure 10.5.2 Some contours of f .

10.5.2 Tree Diagrams

Up to this point, we have applied the Chain Rule to situations where we have
a function z of variables x and y, with both x and y depending on another
single quantity t. We may apply the Chain Rule, however, when x and y each
depend on more than one quantity, or when z is a function of more than two
variables. It can be challenging to keep track of all the dependencies among
the variables, and thus a tree diagram can be a useful tool to organize our
work. For example, suppose that z depends on x and y, and x and y both
depend on t. We may represent these relationships using the tree diagram
shown at left Figure 10.5.3. We place the dependent variable at the top of the
tree and connect it to the variables on which it depends one level below. We
then connect each of those variables to the variable on which each depends.

z z

∂z ∂z
∂x ∂y

x y x y

dx dy
dt dt

t t t t
Figure 10.5.3 A tree diagram illustrating dependencies.
To represent the Chain Rule, we label every edge of the diagram with the
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 154

appropriate derivative or partial derivative, as seen at right in Figure 10.5.3.

To calculate an overall derivative according to the Chain Rule, we construct
the product of the derivatives along all paths connecting the variables and then
add all of these products. For example, the diagram at right in Figure 10.5.3
illustrates the Chain Rule
dz ∂z dx ∂z dy
= + .
dt ∂x dt ∂y dt
Activity 10.5.3
a. Figure 10.5.4 shows the tree diagram we construct when (a) z depends
on w, x, and y, (b) w, x, and y each depend on u and v, and (c) u and
v depend on t.
z

w x y

u v u v u v

t t t t t t

Figure 10.5.4 Three levels of dependencies

i. Label the edges with the appropriate derivatives.

dz
ii. Use the Chain Rule to write dt .

b. Suppose that z = x2 − 2xy 2 and that

x = r cos(θ)
y = r sin(θ).

i. Construct a tree diagram representing the dependencies of z on x

and y and x and y on r and θ.
∂z
ii. Use the tree diagram to find ∂r .
iii. Now suppose that r = 3 and θ = π/6. Find the values of x and y
that correspond to these given values of r and θ, and then use the
Chain Rule to find the value of the partial derivative ∂z
∂θ |(3, 6 ) .
π

10.5.3 Summary
• The Chain Rule is a tool for differentiating a composite for functions. In
its simplest form, it says that if f (x, y) is a function of two variables and
x(t) and y(t) depend on t, then
df ∂f dx ∂f dy
= + .
dt ∂x dt ∂y dt
• A tree diagram can be used to represent the dependence of variables on
other variables. By following the links in the tree diagram, we can form
chains of partial derivatives or derivatives that can be combined to give
a desired partial derivative.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 155

10.5.4 Exercises
dz
1. Use the chain rule to find dt , where

z = x2 y + xy 2 , x = 1 − t5 , y = 3 − t2

First the pieces:

∂z
∂x =
∂z
∂y =
dx
dt =
dy
dt =
End result (in terms of just t):
dz
dt =
Answer 1. 2xy + y 2
Answer 2. x2 + 2xy
Answer 3. (−5)t4
Answer 4. (−2)t1



2 
Answer 5. 2 1 + (−1) t5 3 + (−1) t2 + 3 + (−1) t2 (−1) · 5t4 +
 



2
2 1 + (−1) t5 3 + (−1) t2 + 1 + (−1) t5 (−1) · 2t1
∂z ∂z
2. Use the chain rule to find ∂s and ∂t , where

3s
z = exy tan y, x = 4s + 5t, y =
3t
First the pieces:
∂z ∂z
∂x = ∂y =

∂x ∂x
∂s = ∂t =
∂y ∂y
∂s = ∂t =

And putting it all together:

∂z ∂z ∂x ∂z ∂y ∂z ∂z ∂x ∂z ∂y
∂s = ∂x ∂s + ∂y ∂s and ∂t = ∂x ∂t + ∂y ∂t
Answer 1. y tan(y) exp(xy)
Answer 2. exp(xy) sec2 (y) + tan(y) x exp(xy)
Answer 3. 4
Answer 4. 5
3
Answer 5. 3t
−3s
Answer 6. 3t2
x y
3. Suppose w = + , where
y z
x = e5t , y = 2 + sin (3t), and z = 2 + cos (6t).
A) Use the chain rule to find dw dt as a function of x, y, z, and t. Do
not rewrite x, y, and z in terms of t, and do not rewrite e5t as x.
dw
dt =
Note: You may want to use exp() for the exponential function. Your
answer should be an expression in x, y, z, and t; e.g. “3x - 4y”
B) Use part A to evaluate dw dt when t = 0.
 
5 exp(5t) −y
Answer 1. y + −x 1
y 2 + z · 3 cos(3t) + z 2 (−6) sin(6t)
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 156

Answer 2. 2.75
4. If z = (x + y) ey and x = u2 + v 2 and y = u2 − v 2 , find the following
partial derivatives using the chain rule. Enter your answers as functions
of u and v.
∂z
=
∂u
∂z
=
∂v
2 2 2 2
Answer 1. 2 · 2ueu −v + 2u2 eu −v · 2u
2 2
Answer 2. −2u2 eu −v · 2v
5. If
z = sin x2 + y 2 ,


x = v cos(u) , y = v sin(u) ,
find ∂z/∂u and ∂z/∂v. The variables are restricted to domains on which
the functions are defined.
∂z/∂u =
∂z/∂v =
Answer 1. 0


Answer 2. cos v 2 · 2v
6. Let z = g(u, v) and u(r, s), v(r, s). How many terms are there in the
expression for ∂z/∂r?
terms
Answer. 2
7. Let W (s, t) = F (u(s, t), v(s, t)) where

u(1, 0) = −4, us (1, 0) = −7, ut (1, 0) = −4

v(1, 0) = 8, vs (1, 0) = −8, vt (1, 0) = 3

Fu (−4, 8) = 1, Fv (−4, 8) = 5
Ws (1, 0) = Wt (1, 0) =
Answer 1. −47
Answer 2. 11
8. The radius of a right circular cone is increasing at a rate of 3 inches per
second and its height is decreasing at a rate of 5 inches per second. At
what rate is the volume of the cone changing when the radius is 40 inches
and the height is 40 inches?
cubic inches per second
Answer. 1675.51608191456
9. In a simple electric circuit, Ohm’s law states that V = IR, where V is
the voltage in volts, I is the current in amperes, and R is the resistance
in ohms. Assume that, as the battery wears out, the voltage decreases
at 0.03 volts per second and, as the resistor heats up, the resistance is
increasing at 0.02 ohms per second. When the resistance is 100 ohms and
the current is 0.02 amperes, at what rate is the current changing?

amperes per second

Answer. −0.000304
10. Suppose z = x2 sin y, x = 2s2 + 1t2 , y = −6st.
A. Use the chain rule to find ∂z ∂z
∂s and ∂t as functions of x, y, s and t.
∂z
∂s =
∂z
∂t =
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 157

∂z ∂z
B. Find the numerical values of ∂s and ∂t when (s, t) = (−3, 3).
∂z
∂s (−3, 3) =
∂z
∂t (−3, 3) =
Answer 1. 8sx sin(y) + (−6)tx2 cos(y)
Answer 2. 4tx sin(y) + (−6)sx2 cos(y)
Answer 3. 11244.2989304861
Answer 4. −11063.2512786582
11. Find the indicated derivative. In each case, state the version of the Chain
Rule that you are using.
df
a. dt , if f (x, y) = 2x2 y, x = cos(t), and y = ln(t).
∂f 2z+1
b. ∂w , if f (x, y) = 2x2 y, x = w + z 2 , and y = w
∂f
c. if f (x, y, z) = 2x2 y + z 3 , x = u − v + 2w, y = w2v − u3 , and
∂v ,
z = u2 − v
12. Let z = u2 − v 2 and suppose that

u = ex cos(y)
v = ex sin(y)

a. Find the values of u and v that correspond to x = 0 and y = 2π/3.

b. Use the Chain Rule to find the general partial derivatives
∂z ∂z
and
∂x ∂y
∂z 
 ∂z

and then determine both ∂x (0, 2π )
and ∂y

(0, 2π
.
3 3 )

13. Suppose that T = x2 + y 2 − 2z where

x = ρ sin(φ) cos(θ)
y = ρ sin(φ) sin(θ)
z = ρ cos(φ)

a. Construct a tree diagram representing the dependencies among the

variables.
b. Apply the chain rule to find the partial derivatives
∂T ∂T ∂T
, , and .
∂ρ ∂φ ∂θ
14. Suppose that the temperature on a metal plate is given by the function
T with
T (x, y) = 100 − (x2 + 4y 2 ),
where the temperature is measured in degrees Fahrenheit and x and y are
each measured in feet. Now suppose that an ant is walking on the metal
plate in such a way that it walks in a straight line from the point (1, 4) to
the point (5, 6).
a. Find parametric equations (x(t), y(t)) for the ant’s coordinates as it
walks the line from (1, 4) to (5, 6).
dx dy
b. What can you say about dt and dt for every value of t?
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 158

c. Determine the instantaneous rate of change in temperature with

respect to t that the ant is experiencing at the moment it is halfway
from (1, 4) to (5, 6), using your parametric equations for x and y.
Include units on your answer.
15. There are several proposed formulas to approximate the surface area of
the human body. One model1 uses the formula

A(h, w) = 0.0072h0.725 w0.425 ,

where A is the surface area in square meters, h is the height in centimeters,

and w is the weight in kilograms.
Since a person’s height h and weight w change over time, h and w are
functions of time t. Let us think about what is happening to a child whose
height is 60 centimeters and weight is 9 kilograms. Suppose, furthermore,
that h is increasing at an instantaneous rate of 20 centimeters per year
and w is increasing at an instantaneous rate of 5 kg per year.
Determine the instantaneous rate at which the child’s surface area is
changing at this point in time.
16. Let z = f (x, y) = 50 − (x + 1)2 − (y + 3)2 and z = h(x, y) = 24 − 2x − 6y.
Suppose a person is walking on the surface z = f (x, y) in such a way
that she walks the curve which is the intersection of f and h.
a. Show that x(t) = 4 cos(t) and y(t) = 4 sin(t) is a parameterization
of the “shadow” in the xy-plane of the curve that is the intersection
of the graphs of f and h.

b. Use the parameterization from part (a) to find the instantaneous

rate at which her height is changing with respect to time at the
instant t = 2π/3.
17. The voltage V (in volts) across a circuit is given by Ohm’s Law: V = IR,
where I is the current (in amps) in the circuit and R is the resistance (in
ohms). Suppose we connect two resistors with resistances R1 and R2 in
parallel as shown in Figure 10.5.5. The total resistance R in the circuit is
then given by
1 1 1
= + .
R R1 R2
1 DuBois D, DuBois DF. A formula to estimate the approximate surface area if height

and weight be known. Arch Int Med 1916;17:863-71.

CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 159

a. Assume that the current, I,

and the resistances, R1 and R2 ,
are changing over time, t. Use
the Chain Rule to write a for-
mula for dV R1 R2
dt .

b. Suppose that, at some particu-

lar point in time, we measure + −
the current to be 3 amps and
that the current is increasing
1
at 10 amps per second, whileFigure 10.5.5 Resistors in parallel.
resistance R1 is 2 ohms and de-
creasing at the rate of 0.2 ohms
per second and R2 is 1 ohm
and increasing at the rate of 0.5
ohms per second. At what rate
is the voltage changing at this
point in time?

10.6 Directional Derivatives and the Gradient

Motivating Questions
• The partial derivatives of a function f tell us the rate of change of f in
the direction of the coordinate axes. How can we measure the rate of
change of f in other directions?
• What is the gradient of a function and what does it tell us?
The partial derivatives of a function tell us the instantaneous rate at which
the function changes as we hold all but one independent variable constant and
allow the remaining independent variable to change. It is natural to wonder
how we can measure the rate at which a function changes in directions other
than parallel to a coordinate axes. In what follows, we investigate this question,
and see how the rate of change in any given direction is connected to the rates
of change given by the standard partial derivatives.
Preview Activity 10.6.1 Let’s consider the function f defined by
1
f (x, y) = 30 − x2 − y 2 ,
2
and suppose that f measures the temperature, in degrees Celsius, at a given
point in the plane, where x and y are measured in feet. Assume that the
positive x-axis points due east, while the positive y-axis points due north. A
contour plot of f is shown in Figure 10.6.1
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 160

6 y

1
x
1 2 3 4 5 6

Figure 10.6.1 A contour plot of f (x, y) = 30 − x2 − 21 y 2 .

a. Suppose that a person is walking due east, and thus parallel to the x-axis.
At what instantaneous rate is the temperature changing with respect to
x at the moment the walker passes the point (2, 1)? What are the units
on this rate of change?
b. Next, determine the instantaneous rate of change of temperature with
respect to distance at the point (2, 1) if the person is instead walking due
north. Again, include units on your result.

c. Now, rather than walking due east or due north, let’s suppose that the
person is walking with velocity given by the vector v = h3, 4i, where time
is measured in seconds. Note that the person’s speed is thus |v| = 5 feet
per second. Find parametric equations for the person’s path; that is,
parameterize the line through (2, 1) using the direction vector v = h3, 4i.
Let x(t) denote the x-coordinate of the line, and y(t) its y-coordinate.
Make sure your parameterization places the walker at the point (2, 1)
when t = 0.
d. With the parameterization in (c), we can now view the temperature f
as not only a function of x and y, but also of time, t. Hence, use the
chain rule to determine the value of df
dt t=0 . What are the units on your

answer? What is the practical meaning of this result?

10.6.1 Directional Derivatives

Given a function z = f (x, y), the partial derivative fx (x0 , y0 ) measures the
instantaneous rate of change of f as only the x variable changes; likewise,
fy (x0 , y0 ) measures the rate of change of f at (x0 , y0 ) as only y changes. Note
particularly that fx (x0 , y0 ) is measured in “units of f per unit of change in x,”
and that the units on fy (x0 , y0 ) are similar.
In Preview Activity 10.6.1, we saw how we could measure the rate of change
of f in a situation where both x and y were changing; in that activity, however,
we found that this rate of change was measured in “units of f per unit of time.”
In a given unit of time, we may move more than one unit of distance. In fact,
in Preview Activity 10.6.1, in each unit increase in time we move a distance of
|v| = 5 feet. To generalize the notion of partial derivatives to any direction of
our choice, we instead want to have a rate of change whose units are “units of
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 161

f per unit of distance in the given direction.”

In this light, in order to formally define the derivative in a particular direc-
tion of motion, we want to represent the change in f for a given unit change in
the direction of motion. We can represent this unit change in direction with a
unit vector, say u = hu1 , u2 i. If we move a distance h in the direction of u from
a fixed point (x0 , y0 ), we then arrive at the new point (x0 + u1 h, y0 + u2 h). It
now follows that the slope of the secant line to the curve on the surface through
(x0 , y0 ) in the direction of u through the points (x0 , y0 ) and (x0 +u1 h, y0 +u2 h)
is
f (x0 + u1 h, y0 + u2 h) − f (x0 , y0 )
msec = . (10.6.1)
h
To get the instantaneous rate of change of f in the direction u = hu1 , u2 i,
we must take the limit of the quantity in Equation (10.6.1) as h → 0. Doing
so results in the formal definition of the directional derivative.
Definition 10.6.2 Let f = f (x, y) be given. The derivative of f at the
point (x, y) in the direction of the unit vector u = hu1 , u2 i is denoted
Du f (x, y) and is given by

f (x + u1 h, y + u2 h) − f (x, y)
Du f (x, y) = lim (10.6.2)
h→0 h
for those values of x and y for which the limit exists. ♦
The quantity Du f (x, y) is called a directional derivative. When we evaluate
the directional derivative Du f (x, y) at a point (x0 , y0 ), the result Du f (x0 , y0 )
tells us the instantaneous rate at which f changes at (x0 , y0 ) per unit increase
in the direction of the vector u. In addition, the quantity Du f (x0 , y0 ) tells us
the slope of the line tangent to the surface in the direction of u at the point
(x0 , y0 , f (x0 , y0 )).

10.6.2 Computing the Directional Derivative

In a similar way to how we developed shortcut rules for standard derivatives
in single variable calculus, and for partial derivatives in multivariable calculus,
we can also find a way to evaluate directional derivatives without resorting to
the limit definition found in Equation (10.6.2). We do so using a very similar
approach to our work in Preview Activity 10.6.1.
Suppose we consider the situation where we are interested in the instanta-
neous rate of change of f at a point (x0 , y0 ) in the direction u = hu1 , u2 i, where
u is a unit vector. The variables x and y are therefore changing according to
the parameterization

x = x0 + u1 t and y = y0 + u2 t.
dy
Observe that dx dt = u1 and dt = u2 for all values of t. Since u is a unit
vector, it follows that a point moving along this line moves one unit of distance
per one unit of time; that is, each single unit of time corresponds to movement
of a single unit of distance in that direction. This observation allows us to
use the Chain Rule to calculate the directional derivative, which measures the
instantaneous rate of change of f with respect to change in the direction u.
In particular, by the Chain Rule, it follows that
 
dx  dy 
Du f (x0 , y0 ) = fx (x0 , y0 ) +f y (x 0 , y0 )
dt (x0 ,y0 ) dt (x0 ,y0 )
= fx (x0 , y0 )u1 + fy (x0 , y0 )u2 .
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 162

This now allows us to compute the directional derivative at an arbitrary

point according to the following formula.

Calculating a directional derivative.

Given a differentiable function f = f (x, y) and a unit vector u =

hu1 , u2 i, we may compute Du f (x, y) by

Du f (x, y) = fx (x, y)u1 + fy (x, y)u2 . (10.6.3)

Note well: To use Equation (10.6.3), we must have a unit vector u =

hu1 , u2 i in the direction of motion. In the event that we have a direction
prescribed by a non-unit vector, we must first scale the vector to have length
1.
Activity 10.6.2 Let f (x, y) = 3xy − x2 y 3 .
a. Determine fx (x, y) and fy (x, y).
b. Use Equation (10.6.3) to determine Di f (x, y) and Dj f (x, y). What fa-
miliar function is Di f ? What familiar function is Dj f ? (Recall that i is
the unit vector in the positive x-direction and j is the unit vector in the
positive y-direction.)
c. Use Equation (10.6.3) to find the derivative of f in the direction of the
vector v = h2, 3i at the point (1, −1). Remember that a unit direction
vector is needed.

10.6.3 The Gradient

Via the Chain Rule, we have seen that for a given function f = f (x, y), its
instantaneous rate of change in the direction of a unit vector u = hu1 , u2 i is
given by
Du f (x0 , y0 ) = fx (x0 , y0 )u1 + fy (x0 , y0 )u2 . (10.6.4)
Recalling that the dot product of two vectors v = hv1 , v2 i and u = hu1 , u2 i
is computed by
v · u = v1 u1 + v2 u2 ,
we see that we may recast Equation (10.6.4) in a way that has geometric
meaning. In particular, we see that Du f (x0 , y0 ) is the dot product of the
vector hfx (x0 , y0 ), fy (x0 , y0 )i and the vector u.
We call this vector formed by the partial derivatives of f the gradient of f
and denote it
∇f (x0 , y0 ) = hfx (x0 , y0 ), fy (x0 , y0 )i .
We read ∇f as “the gradient of f ,” “grad f ” or “del f ”.1 Notice that ∇f
varies from point to point, and also provides an alternate formulation of the
directional derivative.
The directional derivative and the gradient.

Given a differentiable function f = f (x, y) and a unit vector u =

hu1 , u2 i, we may compute Du f (x, y) by

Du f (x, y) = ∇f (x, y) · u. (10.6.5)

1 The symbol ∇ is called nabla, which comes from a Greek word for a certain type of harp

that has a similar shape.

CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 163

In the following activity, we investigate some of what the gradient tells us

about the behavior of a function f .
Activity 10.6.3 Let’s consider the function f defined by f (x, y) = x2 − y 2 .
Some contours for this function are shown in Figure 10.6.3.
4 y
3
2
1
x
-4 -3 -2 -1 1 2 3 4
-1 3 6 9
0
-2 -3
-6
-3 -9
-4

Figure 10.6.3 Contours of f (x, y) = x2 − y 2 .

a. Find the gradient ∇f (x, y).

b. For each of the following points (x0 , y0 ), evaluate the gradient ∇f (x0 , y0 )
and sketch the gradient vector with its tail at (x0 , y0 ). Some of the vectors
are too long to fit onto the plot, but we’d like to draw them to scale; to
do so, scale each vector by a factor of 1/4.

• (x0 , y0 ) = (2, 0)
• (x0 , y0 ) = (0, 2)
• (x0 , y0 ) = (2, 2)
• (x0 , y0 ) = (2, 1)
• (x0 , y0 ) = (−3, 2)
• (x0 , y0 ) = (−2, −4)
• (x0 , y0 ) = (0, 0)
c. What do you notice about the relationship between the gradient at
(x0 , y0 ) and the contour passing through that point?

d. Does f increase or decrease in the direction of ∇f (x0 , y0 )? Provide a

justification for your response.
As a vector, ∇f (x0 , y0 ) defines a direction and a length. As we will soon
see, both of these convey important information about the behavior of f near
(x0 , y0 ).

10.6.4 The Direction of the Gradient

Remember that the dot product also conveys information about the angle
between the two vectors. If θ is the angle between ∇f (x0 , y0 ) and u (where u
is a unit vector), then we also have that

Du f (x0 , y0 ) = ∇f (x0 , y0 ) · u = |∇f (x0 , y0 )||u| cos(θ).

CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 164

In particular, when θ is a right angle, as shown on the left of Figure 10.6.4,

then Du f (x0 , y0 ) = 0, because cos(θ) = 0. Since the value of the directional
derivative is 0, this means that f is unchanging in this direction, and hence
u must be tangent to the contour of f that passes through (x0 , y0 ). In other
words, ∇f (x0 , y0 ) is orthogonal to the contour through (x0 , y0 ). This shows
that the gradient vector at a given point is always perpendicular to the con-
tour passing through the point, confirming that what we saw in part (c) of
Activity 10.6.3 holds in general.

y y y
∇f (x0 , y0 ) ∇f (x0 , y0 ) ∇f (x0 , y0 )

u
θ θ θ
u
(x0 , y0 ) (x0 , y0 ) u (x0 , y0 )
x x x

Figure 10.6.4 The sign of Du f (x0 , y0 ) is determined by θ.

Moreover, when θ is an acute angle, it follows that cos(θ) > 0 so since
Du f (x0 , y0 ) = |∇f (x0 , y0 )||u| cos(θ),
and therefore Du f (x0 , y0 ) > 0, as shown in the middle image in Figure 10.6.4.
This means that f is increasing in any direction where θ is acute. In a similar
way, when θ is an obtuse angle, then cos(θ) < 0 so Du f (x0 , y0 ) < 0, as seen
on the right in Figure 10.6.4. This means that f is decreasing in any direction
for which θ is obtuse.
Finally, as we can see in the following activity, we may also use the gradient
to determine the directions in which the function is increasing and decreasing
most rapidly.
Activity 10.6.4 In this activity we investigate how the gradient is related
to the directions of greatest increase and decrease of a function. Let f be a
differentiable function and u a unit vector.
a. Let θ be the angle between ∇f (x0 , y0 ) and u. Use the relationship be-
tween the dot product and the angle between two vectors to explain why
Du f (x0 , y0 ) = |hfx (x0 , y0 ), fy (x0 , y0 )i| cos(θ). (10.6.6)

b. At the point (x0 , y0 ), the only quantity in Equation (10.6.6) that can
change is θ (which determines the direction u of travel). Explain why
θ = 0 makes the quantity
|hfx (x0 , y0 ), fy (x0 , y0 )i| cos(θ)
as large as possible.
c. When θ = 0, in what direction does the unit vector u point relative to
∇f (x0 , y0 )? Why? What does this tell us about the direction of greatest
increase of f at the point (x0 , y0 )?
d. In what direction, relative to ∇f (x0 , y0 ), does f decrease most rapidly
at the point (x0 , y0 )?
e. State the unit vectors u and v (in terms of ∇f (x0 , y0 )) that provide the
directions of greatest increase and decrease for the function f at the point
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 165

(x0 , y0 ). What important assumption must we make regarding ∇f (x0 , y0 )

in order for these vectors to exist?

10.6.5 The Length of the Gradient

Having established in Activity 10.6.4 that the direction in which a function
increases most rapidly at a point (x0 , y0 ) is the unit vector u in the direction
of the gradient, (that is, u = |∇f (x10 ,y0 )| ∇f (x0 , y0 ), provided that ∇f (x0 , y0 ) 6=
0), it is also natural to ask, “in the direction of greatest increase for f at
(x0 , y0 ), what is the value of the rate of increase?” In this situation, we are
asking for the value of Du f (x0 , y0 ) where u = |∇f (x10 ,y0 )| ∇f (x0 , y0 ).
Using the now familiar way to compute the directional derivative, we see
that  
1
Du f (x0 , y0 ) = ∇f (x0 , y0 ) · ∇f (x0 , y0 ) .
|∇f (x0 , y0 )|
Next, we recall two important facts about the dot product: (i) w · (cv) =
c(w · v) for any scalar c, and (ii) w · w = |w|2 . Applying these properties to
the most recent equation involving the directional derivative, we find that
1 1
Du f (x0 , y0 ) = (∇f (x0 , y0 )·∇f (x0 , y0 )) = |∇f (x0 , y0 )|2 .
|∇f (x0 , y0 )| |∇f (x0 , y0 )|

Finally, since ∇f (x0 , y0 ) is a nonzero vector, its length |∇f (x0 , y0 )| is a

nonzero scalar, and thus we can simplify the preceding equation to establish
that
Du f (x0 , y0 ) = |∇f (x0 , y0 )|.
We summarize our most recent work by stating two important facts about
the gradient.

Important facts about the gradient.

Let f be a differentiable function and (x0 , y0 ) a point for which ∇f (x0 , y0 ) 6=

0. Then ∇f (x0 , y0 ) points in the direction of greatest increase of f at
(x0 , y0 ), and the instantaneous rate of change of f in that direction is
the length of the gradient vector. That is, if u = |∇f (x10 ,y0 )| ∇f (x0 , y0 ),
then u is a unit vector in the direction of greatest increase of f at
(x0 , y0 ), and Du f (x0 , y0 ) = |∇f (x0 , y0 )|.

Activity 10.6.5 Consider the function f defined by f (x, y) = −x + 2xy − y.

a. Find the gradient ∇f (1, 2) and sketch it on Figure 10.6.5.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 166

4 y

x
1 2 3 4

Figure 10.6.5 A plot for the gradient ∇f (1, 2).

D E
b. Sketch the unit vector z = − √12 , − √12 on Figure 10.6.5 with its tail at
(1, 2). Now find the directional derivative Dz f (1, 2).
c. What is the slope of the graph of f in the direction z? What does the
sign of the directional derivative tell you?
d. Consider the vector v = h2, −1i and sketch v on Figure 10.6.5 with its
tail at (1, 2). Find a unit vector w pointing in the same direction of
v. Without computing Dw f (1, 2), what do you know about the sign of
this directional derivative? Now verify your observation by computing
Dw f (1, 2).
e. In which direction (that is, for what unit vector u) is Du f (1, 2) the
greatest? What is the slope of the graph in this direction?

f. Corresponding, in which direction is Du f (1, 2) least? What is the slope

of the graph in this direction?
g. Sketch two unit vectors u for which Du f (1, 2) = 0 and then find compo-
nent representations of these vectors.

h. Suppose you are standing at the point (3, 3). In which direction should
you move to cause f to increase as rapidly as possible? At what rate
does f increase in this direction?

10.6.6 Applications
The gradient finds many natural applications. For example, situations often
arise — for instance, constructing a road through the mountains or planning
the flow of water across a landscape — where we are interested in knowing the
direction in which a function is increasing or decreasing most rapidly.
For example, consider a two-dimensional version of how a heat-seeking mis-
sile might work.(This application is borrowed from United States Air Force
Academy Department of Mathematical Sciences.) Suppose that the tempera-
ture surrounding a fighter jet can be modeled by the function T defined by
100
T (x, y) = ,
1 + (x − 5)2 + 4(y − 2.5)2
where (x, y) is a point in the plane of the fighter jet and T (x, y) is measured
in degrees Celsius. Some contours and gradients ∇T are shown on the left in
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 167

Figure 10.6.6.
5 y 5 y

4 4

3 3

2 2

1 1
x x
1 2 3 4 5 6 7 8 9 10 1 2 3 4 5 6 7 8 9 10

Figure 10.6.6 Contours and gradient for T (x, y) and the missile’s path.
A heat-seeking missile will always travel in the direction in which the tem-
perature increases most rapidly; that is, it will always travel in the direction
of the gradient ∇T . If a missile is fired from the point (2, 4), then its path will
be that shown on the right in Figure 10.6.6.
In the final activity of this section, we consider several questions related to
this context of a heat-seeking missile, and foreshadow some upcoming work in
Section 10.7.
Activity 10.6.6
a. The temperature T (x, y) has its maximum value at the fighter jet’s loca-
tion. State the fighter jet’s location and explain how Figure 10.6.6 tells
you this.
b. Determine ∇T at the fighter jet’s location and give a justification for
your response.
c. Suppose that a different function f has a local maximum value at (x0 , y0 ).
Sketch the behavior of some possible contours near this point. What is
∇f (x0 , y0 )? (Hint: What is the direction of greatest increase in f at the
local maximum?)
d. Suppose that a function g has a local minimum value at (x0 , y0 ). Sketch
the behavior of some possible contours near this point. What is ∇g(x0 , y0 )?

e. If a function g has a local minimum at (x0 , y0 ), what is the direction of

greatest increase of g at (x0 , y0 )?

10.6.7 Summary
• The directional derivative of f at the point (x, y) in the direction of the
unit vector u = hu1 , u2 i is
f (x + u1 h, y + u2 h) − f (x, y)
Du f (x, y) = lim
h→0 h
for those values of x and y for which the limit exists. In addition,
Du f (x, y) measures the slope of the graph of f when we move in the
direction u. Alternatively, Du f (x0 , y0 ) measures the instantaneous rate
of change of f in the direction u at (x0 , y0 ).
• The gradient of a function f = f (x, y) at a point (x0 , y0 ) is the vector
∇f (x0 , y0 ) = hfx (x0 , y0 ), fy (x0 , y0 )i .

• The directional derivative in the direction u may be computed by

Du f (x0 , y0 ) = ∇f (x0 , y0 ) · u.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 168

• At any point where the gradient is nonzero, gradient is orthogonal to

the contour through that point and points in the direction in which f
increases most rapidly; moreover, the slope of f in this direction equals
the length of the gradient |∇f (x0 , y0 )|. Similarly, the opposite of the
gradient points in the direction of greatest decrease, and that rate of
decrease is the opposite of the length of the gradient.

10.6.8 Exercises
1. Consider the function f (x, y, z) = xy + yz 2 + xz 3 .
Find the gradient of f :
h , ,
i
Find the gradient of f at the point (2, 3, 5).
h , ,
i
Find the rate √
of change√of the function
√ f at the point (2, 3,5) in the
direction u = h4/ 66, −5/ 66, −5/ 66i.
Answer 1. y + z 3
Answer 2. x + z 2
Answer 3. 2yz + 3xz 2
Answer 4. 128
Answer 5. 27
Answer 6. 180
Answer 7. −64.376849782191
2. If f (x, y) = 3x2 − 1y 2 , find the value of the directional derivative at the
point (−1, −4) in the direction given by the angle θ = 2π 6 .
Answer. 3.92820323027551
3. Find the directional derivative of f (x, y, z) = 4xy + z 2 at the point
(−5, 3, −2) in the direction of the maximum rate of change of f .
f<B><I>u</B></I> (−5, 3, −2) = D<B><I>u</B></I> f (−5, 3, −2) =

Answer. 23.6643
100
4. The temperature at any point in the plane is given by T (x, y) = .
x2 + y2 + 3
(a) What shape are the level curves of T ?

 hyperbolas
 circles

 lines
 ellipses
 parabolas

 none of the above

(b) At what point on the plane is it hottest?

What is the maximum temperature?
(c) Find the direction of the greatest increase in temperature at the
point (−3, −3).
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 169

What is the value of this maximum rate of change, that is, the maxi-
mum value of the directional derivative at (−3, −3)?
(d) Find the direction of the greatest decrease in temperature at the
point (−3, −3).
What is the value of this most negative rate of change, that is, the
minimum value of the directional derivative at (−3, −3)?
Answer 1. B
Answer 2. (0, 0)
Answer 3. 33.3333
Answer 4. h1.36054, 1.36054i
Answer 5. 1.9241
Answer 6. h−1.36054, −1.36054i
2 2 2
5. The temperature at a point (x,y,z) is given by T (x, y, z) = 200e−x −y /4−z /9 ,
where T is measured in degrees Celsius and x,y, and z in meters. There
are lots of places to make silly errors in this problem; just try to keep
track of what needs to be a unit vector.
Find the rate of change of the temperature at the point (-1, 1, 2) in
the direction toward the point (-1, 3, -3).
In which direction (unit vector) does the temperature increase the
fastest at (-1, 1, 2)?
h , ,
i
What is the maximum rate of increase of T at (-1, 1, 2)?
Answer 1. 8.33860305193074
Answer 2. 0.948354065592898
Answer 3. −0.237088516398225
Answer 4. −0.210745347909533
Answer 5. 77.4821403113886
6. If f (x, y, z) = 2zy 2 , then the gradient at the point (2, 2, 4) is
∇f (2, 2, 4) =
Answer. h0, 32, 8i
7. The concentration of salt in a fluid at (x, y, z) is given by F (x, y, z) =
2x2 + 3y 4 + 2x2 z 2 mg/cm3 . You are at the point (−1, 1, −1).
(a) In which direction should you move if you want the concentration
to increase the fastest?
direction:
(Give your answer as a vector.)
(b) You start to move in the direction you found in part (a) at a speed
of 5 cm/sec. How fast is the concentration changing?
rate of change =
Answer 1. −8 ĩ + 12 j̃ − 4 k̃
p
Answer 2. 5 −8 · (−8) + 12 · 12 + −4 · (−4)
8. At a certain point on a heated metal plate, the greatest rate of tempera-
ture increase, 4 degrees Celsius per meter, is toward the northeast. If an
object at this point moves directly north, at what rate is the temperature
increasing?

degrees Celsius per meter

Answer. √4
2
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 170

9. Suppose that you are climbing a hill whose shape is given by z = 902 −
0.07x2 − 0.1y 2 , and that you are at the point (40, 70, 300).
In which direction (unit vector) should you proceed initially in order
to reach the top of the hill fastest?
h ,
i
If you climb in that direction, at what angle above the horizontal will
you be climbing initially (radian measure)?
Answer 1. −0.371390676354104
Answer 2. −0.928476690885259
Answer 3. 1.5045735390203
10. Are the following statements true or false?
(a) The gradient vector ∇f (a, b) is tangent to the contour of f at (a, b).
(b) f~u (a, b) = ||∇f (a, b)||.
(c) f~u (a, b) is parallel to ~u.

(d) If ~u is perpendicular to ∇f (a, b), then f~u (a, b) = h0, 0i.

(e) If ~u is a unit vector, then f~u (a, b) is a vector.
(f) Suppose fx (a, b) and fy (a, b) both exist. Then there is always a
direction in which the rate of change of f at (a, b) is zero.

(g) If f (x, y) has fx (a, b) = 0 and fy (a, b) = 0 at the point (a, b), then
f is constant everywhere.
(h) ∇f (a, b) is a vector in 3-dimensional space.
11. Let E(x, y) = 1+(x−5)2100
+4(y−2.5)2 represent the elevation on a land mass
at location (x, y). Suppose that E, x, and y are all measured in meters.
a. Find Ex (x, y) and Ey (x, y).

b. Let u be a unit vector in the direction of h−4, 3i. Determine Du E(3, 4).
What is the practical meaning of Du E(3, 4) and what are its units?
c. Find the direction of greatest increase in E at the point (3, 4).
d. Find the instantaneous rate of change of E in the direction of great-
est decrease at the point (3, 4). Include units on your answer.

e. At the point (3, 4), find a direction w in which the instantaneous

rate of change of E is 0.
12. Find all directions in which the directional derivative of f (x, y) = ye−xy
is 1 at the point (0, 2).
13. Find, if possible, a function f such that
 
5
∇f = sin(yz), xz cos(yz) + 2y, xy cos(yz) + .
z

If not possible, explain why.

14. Let f (x, y) = x2 + 3y 2 .
a. Find ∇f (x, y) and ∇f (1, 2).

b. Find the direction of greatest increase in f at the point (1, 2). Ex-
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 171

plain. A graph of the surface defined by f is shown at left in Fig-

ure 10.6.7. Illustrate this direction on the surface.
c. A contour diagram of f is shown at right in Figure 10.6.7. Illustrate
your calculation from (b) on this contour diagram.
4 y
z 3
60
2
40 1
x
20 -4 -3 -2 -1 1 2 3 4
-1

-2
-4 -2
0 x
-4 -2 0 2 2
4
y -3
-4

Figure 10.6.7 Left: Graph of f (x, y) = x2 + 3y 2 . Right: Contours.

d. Find a direction w for which the derivative of f in the direction of

w is zero.
15. The properties of the gradient that we have observed for functions of two
variables also hold for functions of more variables. In this problem, we
consider a situation where there are three independent variables. Suppose
that the temperature in a region of space is described by
2
−y 2 −z 2
T (x, y, z) = 100e−x

and that you are standing at the point (1, 2, −1).

a. Find the instantaneous rate of change of the temperature in the
direction of v = h0, 1, 2i at the point (1, 2, −1). Remember that you
should first find a unit vector in the direction of v.
b. In what direction from the point (1, 2, −1) would you move to cause
the temperature to decrease as quickly as possible?
c. How fast does the temperature decrease in this direction?

d. Find a direction in which the temperature does not change at (1, 2, −1).
16. Figure 10.6.8 shows a plot of the gradient ∇f at several points for some
function f = f (x, y).
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 172

Figure 10.6.8 The gradient ∇f .

a. Consider each of the three indicated points, and draw, as best as

you can, the contour through that point.

b. Beginning at each point, draw a curve on which f is continually

decreasing.

10.7 Optimization
Motivating Questions
• How can we find the points at which f (x, y) has a local maximum or
minimum?
• How can we determine whether critical points of f (x, y) are local maxima
or minima?
• How can we find the absolute maximum and minimum of f (x, y) on a
closed and bounded domain?
We learn in single-variable calculus that the derivative is a useful tool for
finding the local maxima and minima of functions, and that these ideas may
often be employed in applied settings. In particular, if a function f , such
as the one shown in Figure 10.7.1 is everywhere differentiable, we know that
the tangent line is horizontal at any point where f has a local maximum or
minimum. This, of course, means that the derivative f 0 is zero at any such
point. Hence, one way that we seek extreme values of a given function is to
first find where the derivative of the function is zero.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 173

y = f (x)

Figure 10.7.1 The graph of y = f (x).

In multivariable calculus, we are often similarly interested in finding the
greatest and/or least value(s) that a function may achieve. Moreover, there
are many applied settings in which a quantity of interest depends on several
different variables. In the following preview activity, we begin to see how
some key ideas in multivariable calculus can help us answer such questions
by thinking about the geometry of the surface generated by a function of two
variables.
Preview Activity 10.7.1 Let z = f (x, y) be a differentiable function, and
suppose that at the point (x0 , y0 ), f achieves a local maximum. That is, the
value of f (x0 , y0 ) is greater than the value of f (x, y) for all (x, y) nearby (x0 , y0 ).
You might find it helpful to sketch a rough picture of a possible function f that
has this property.
a. If we consider the trace given by holding y = y0 constant, then the single-
variable function defined by f (x, y0 ) must have a local maximum at x0 .
What does this say about the value of the partial derivative fx (x0 , y0 )?

b. In the same way, the trace given by holding x = x0 constant has a local
maximum at y = y0 . What does this say about the value of the partial
derivative fy (x0 , y0 )?
c. What may we now conclude about the gradient ∇f (x0 , y0 ) at the local
maximum? How is this consistent with the statement “f increases most
rapidly in the direction ∇f (x0 , y0 )?”
d. How will the tangent plane to the surface z = f (x, y) appear at the point
(x0 , y0 , f (x0 , y0 ))?
e. By first computing the partial derivatives, find any points at which
f (x, y) = 2x − x2 − (y + 2)2 may have a local maximum.

10.7.1 Extrema and Critical Points

One of the important applications of single-variable calculus is the use of deriva-
tives to identify local extremes of functions (that is, local maxima and local
minima). Using the tools we have developed so far, we can naturally extend
the concept of local maxima and minima to several-variable functions.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 174

Definition 10.7.2 Let f be a function of two variables x and y.

• The function f has a local maximum at a point (x0 , y0 ) provided that
f (x, y) ≤ f (x0 , y0 ) for all points (x, y) near (x0 , y0 ). In this situation we
say that f (x0 , y0 ) is a local maximum value.
• The function f has a local minimum at a point (x0 , y0 ) provided that
f (x, y) ≥ f (x0 , y0 ) for all points (x, y) near (x0 , y0 ). In this situation we
say that f (x0 , y0 ) is a local minimum value.
• An absolute maximum point is a point (x0 , y0 ) for which f (x, y) ≤ f (x0 , y0 )
for all points (x, y) in the domain of f . The value of f at an absolute
maximum point is the maximum value of f .

• An absolute minimum point is a point such that f (x, y) ≥ f (x0 , y0 ) for all
points (x, y) in the domain of f . The value of f at an absolute minimum
point is the maximum value of f .
♦
We use the term extremum point to refer to any point (x0 , y0 ) at which f
has a local maximum or minimum. In addition, the function value f (x0 , y0 ) at
an extremum is called an extremal value. Figure 10.7.3 illustrates the graphs
of two functions that have an absolute maximum and minimum, respectively,
at the origin (x0 , y0 ) = (0, 0).

Figure 10.7.3 An absolute maximum and an absolute minimum

In single-variable calculus, we saw that the extrema of a continuous function
f always occur at critical points, values of x where f fails to be differentiable
or where f 0 (x) = 0. Said differently, critical points provide the locations where
extrema of a function may appear. Our work in Preview Activity 10.7.1 sug-
gests that something similar happens with two-variable functions.
Suppose that a continuous function f has an extremum at (x0 , y0 ). In this
case, the trace f (x, y0 ) has an extremum at x0 , which means that x0 is a critical
value of f (x, y0 ). Therefore, either fx (x0 , y0 ) does not exist or fx (x0 , y0 ) = 0.
Similarly, either fy (x0 , y0 ) does not exist or fy (x0 , y0 ) = 0. This implies that
the extrema of a two-variable function occur at points that satisfy the following
definition.
Definition 10.7.4 A critical point (x0 , y0 ) of a function f = f (x, y) is a
point in the domain of f at which fx (x0 , y0 ) = 0 and fy (x0 , y0 ) = 0, or such
that one of fx (x0 , y0 ) or fy (x0 , y0 ) fails to exist. ♦
We can therefore find critical points of a function f by computing partial
derivatives and identifying any values of (x, y) for which one of the partials
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 175

doesn’t exist or for which both partial derivatives are simultaneously zero. For
the latter, note that we have to solve the system of equations

fx (x, y) = 0
fy (x, y) = 0.

Activity 10.7.2 Find the critical points of each of the following functions.
Then, using appropriate technology, plot the graphs of the surfaces near each
critical value and compare the graph to your work.
a. f (x, y) = 2 + x2 + y 2
b. f (x, y) = 2 + x2 − y 2
c. f (x, y) = 2x − x2 − 14 y 2

d. f (x, y) = |x| + |y|

e. f (x, y) = 2xy − 4x + 2y − 3.

10.7.2 Classifying Critical Points: The Second Derivative

Test
While the extrema of a continuous function f always occur at critical points, it
is important to note that not every critical point leads to an extremum. Recall,
for instance, f (x) = x3 from single variable calculus. We know that x0 = 0 is
a critical point since f 0 (x0 ) = 0, but x0 = 0 is neither a local maximum nor a
local minimum of f .
A similar situation may arise in a multivariable setting. Consider the func-
tion f defined by f (x, y) = x2 − y 2 whose graph and contour plot are shown in
Figure 10.7.5. Because ∇f = h2x, −2yi, we see that the origin (x0 , y0 ) = (0, 0)
is a critical point. However, this critical point is neither a local maximum or
minimum; the origin is a local minimum on the trace defined by y = 0, while
the origin is a local maximum on the trace defined by x = 0. We call such
a critical point a saddle point due to the shape of the graph near the critical
point.

x y

Figure 10.7.5 A saddle point.

As in single-variable calculus, we would like to have some sort of test to
help us identify whether a critical point is a local maximum, local maximum,
or neither.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 176

Activity 10.7.3 Recall that the Second Derivative Test for single-variable
functions states that if x0 is a critical point of a function f so that f 0 (x0 ) = 0
and if f 00 (x0 ) exists, then
• if f 00 (x0 ) < 0, x0 is a local maximum,
• if f 00 (x0 ) > 0, x0 is a local minimum, and

• if f 00 (x0 ) = 0, this test yields no information.

Our goal in this activity is to understand a similar test for classifying ex-
treme values of functions of two variables. Consider the following three func-
tions:

f1 (x, y) = 4 − x2 − y 2 , f2 (x, y) = x2 + y 2 , f3 (x, y) = x2 − y 2 .

You can verify that each function has a critical point at the origin (0, 0). You
should check this.

x y

Figure 10.7.6 Three surfaces.

a. The graphs of these three functions are shown in Figure 10.7.6, with
z = 4 − x2 − y 2 at left, z = x2 + y 2 in the middle, and z = x2 − y 2 at
right. Use the graphs to decide if a function has a local maximum, local
minimum, saddle point, or none of the above at the origin.
b. There is no single second derivative of a function of two variables, so we
consider a quantity that combines the second order partial derivatives.
2
Let D = fxx fyy − fxy . Calculate D at the origin for each of the functions
f1 , f2 , and f3 . What difference do you notice between the values of D
when a function has a maximum or minimum value at the origin versus
when a function has a saddle point at the origin?

c. Now consider the cases where D > 0. It is in these cases that a function
has a local maximum or minimum at a point. What is necessary in these
cases is to find a condition that will distinguish between a maximum and
a minimum. In the cases where D > 0 at the origin, evaluate fxx (0, 0).
What value does fxx (0, 0) have when f has a local maximum value at
the origin? When f has a local minimum value at the origin? Explain
why. (Hint: This should look very similar to the Second Derivative Test
for functions of a single variable.) What would happen if we considered
the values of fyy (0, 0) instead?
Activity 10.7.3 provides the basic ideas for the Second Derivative Test for
functions of two variables.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 177

The Second Derivative Test.

Suppose (x0 , y0 ) is a critical point of the function f for which fx (x0 , y0 ) =
0 and fy (x0 , y0 ) = 0. Let D be the quantity defined by

D = fxx (x0 , y0 )fyy (x0 , y0 ) − fxy (x0 , y0 )2 .

• If D > 0 and fxx (x0 , y0 ) < 0, then f has a local maximum at

(x0 , y0 ).
• If D > 0 and fxx (x0 , y0 ) > 0, then f has a local minimum at
(x0 , y0 ).

• If D < 0, then f has a saddle point at (x0 , y0 ).

• If D = 0, then this test yields no information about what happens
at (x0 , y0 ).
The quantity D is called the discriminant of the function f at
(x0 , y0 ).

To properly understand the origin of the Second Derivative Test, we could

introduce a “second-order directional derivative.” If this second-order direc-
tional derivative were negative in every direction, for instance, we could guar-
antee that the critical point is a local maximum. A complete justification of
the Second Derivative Test requires key ideas from linear algebra that are be-
yond the scope of this course, so instead of presenting a detailed explanation,
we will accept this test as stated. In Activity 10.7.4, we apply the test to more
complicated examples.
Activity 10.7.4 Find the critical points of the following functions and use the
Second Derivative Test to classify the critical points.
a. f (x, y) = 3x3 + y 2 − 9x + 4y
2 4
b. f (x, y) = xy + x + y

c. f (x, y) = x3 + y 3 − 3xy.
As we learned in single-variable calculus, finding extremal values of func-
tions can be particularly useful in applied settings. For instance, we can often
use calculus to determine the least expensive way to construct something or to
find the most efficient route between two locations. The same possibility holds
in settings with two or more variables.
Activity 10.7.5 While the quantity of a product demanded by consumers is
often a function of the price of the product, the demand for a product may also
depend on the price of other products. For instance, the demand for blue jeans
at Old Navy may be affected not only by the price of the jeans themselves, but
also by the price of khakis.
Suppose we have two goods whose respective prices are p1 and p2 . The
demand for these goods, q1 and q2 , depend on the prices as

q1 = 150 − 2p1 − p2 (10.7.1)

q2 = 200 − p1 − 3p2 . (10.7.2)

The seller would like to set the prices p1 and p2 in order to maximize
revenue. We will assume that the seller meets the full demand for each product.
Thus, if we let R be the revenue obtained by selling q1 items of the first good
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 178

at price p1 per item and q2 items of the second good at price p2 per item, we
have
R = p1 q1 + p2 q2 .
We can then write the revenue as a function of just the two variables p1
and p2 by using Equations (10.7.1) and (10.7.2), giving us

R(p1 , p2 ) = p1 (150 − 2p1 − p2 ) + p2 (200 − p1 − 3p2 )

= 150p1 + 200p2 − 2p1 p2 − 2p21 − 3p22 .

A graph of R as a function of p1 and p2 is shown in Figure 10.7.7.

4000

2000

-2000

-4000

-6000

-8000

-10000
10 20 30 40 20 10 0
p2 50 60 60 50 40 30 p1

Figure 10.7.7 A revenue function.

a. Find all critical points of the revenue function, R. (Hint: You should
obtain a system of two equations in two unknowns which can be solved
by elimination or substitution.)
b. Apply the Second Derivative Test to determine the type of any critical
point(s).
c. Where should the seller set the prices p1 and p2 to maximize the revenue?

10.7.3 Optimization on a Restricted Domain

The Second Derivative Test helps us classify critical points of a function, but it
does not tell us if the function actually has an absolute maximum or minimum
at each such point. For single-variable functions, the Extreme Value Theorem
told us that a continuous function on a closed interval [a, b] always has both
an absolute maximum and minimum on that interval, and that these absolute
extremes must occur at either an endpoint or at a critical point. Thus, to
find the absolute maximum and minimum, we determine the critical points
in the interval and then evaluate the function at these critical points and at
the endpoints of the interval. A similar approach works for functions of two
variables.
For functions of two variables, closed and bounded regions play the role
that closed intervals did for functions of a single variable. A closed region is a
region that contains its boundary (the unit disk x2 + y 2 ≤ 1 is closed, while its
interior x2 + y 2 < 1 is not, for example), while a bounded region is one that
does not stretch to infinity in any direction. Just as for functions of a single
variable, continuous functions of several variables that are defined on closed,
bounded regions must have absolute maxima and minima in those regions.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 179

The Extreme Value Theorem.

Let f = f (x, y) be a continuous function on a closed and bounded
region R. Then f has an absolute maximum and an absolute minimum
in R.

The absolute extremes must occur at either a critical point in the interior
of R or at a boundary point of R. We therefore must test both possibilities,
as we demonstrate in the following example.
Example 10.7.8 Suppose the temperature T at each point on the circular
plate x2 + y 2 ≤ 1 is given by
T (x, y) = 2x2 + y 2 − y.
The domain R = {(x, y) : x2 + y 2 ≤ 1} is a closed and bounded region, as
shown on the left of Figure 10.7.9, so the Extreme Value Theorem assures us
that T has an absolute maximum and minimum on the plate. The graph of
T over its domain R is shown in Figure 10.7.9. We will find the hottest and
coldest points on the plate.
1.5 y
z
2
1.0

0.5
1
x
-1.5 -1.0 -0.5 0.5 1.0 1.5
-0.5 -1
0 0 x
y 1
-1.0
-1

-1.5
Figure 10.7.9 Domain of the temperature T (x, y) = 2x2 + y 2 − y and its
graph.
If the absolute maximum or minimum occurs inside the disk, it will be at
a critical point so we begin by looking for critical points inside the disk. To do
this, notice that critical points are given by the conditions Tx = 4x = 0 and
Ty = 2y − 1 = 0. This means that there is one critical point of the function at
the point (x0 , y0 ) = (0, 1/2), which lies inside the disk.
We now find the hottest and coldest points on the boundary of the disk,
which is the circle of radius 1. As we have seen, the points on the unit circle
can be parametrized as
x(t) = cos(t), y(t) = sin(t),
where 0 ≤ t ≤ 2π. The temperature at a point on the circle is then described
by
T (x(t), y(t)) = 2 cos2 (t) + sin2 (t) − sin(t).
To find the hottest and coldest points on the boundary, we look for the
critical points of this single-variable function on the interval 0 ≤ t ≤ 2π. We
have
dT
= −4 cos(t) sin(t) + 2 cos(t) sin(t) − cos(t)
dt
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 180

= −2 cos(t) sin(t) − cos(t) = cos(t)(−2 sin(t) − 1)

= 0.

This shows that we have critical points when cos(t) = 0 or sin(t) = −1/2.
This occurs when t = π/2, 3π/2, 7π/6, and 11π/6. Since we have x(t) = cos(t)
and y(t) = sin(t), the corresponding points are
π 3π
• (x, y) =   t = 2,
(0, 1) when • (x, y) = 
(0, −1) when
 t= 2 ,
√ √
• (x, y) = 23 , − 12 when t = 11π
6 . • (x, y) = − 23 , − 21 when t = 7π
6 .

These are the critical points of T on the boundary and so this collection of
points includes the hottest and coldest points on the boundary.
We now have a list of candidates for the hottest and coldest points: the
critical point in the interior of the disk and the critical points on the boundary.
We find the hottest and coldest points by evaluating the temperature at each
of these points, and find that
• T 0, 12 = −14 ,


• T (0,
 1) = 0,  • T (0, −1) = 2,
√ √
3 1 9 3 1 9
• T − 2 , −2 = 4, • T − 2 , −2 = 4.
2 2 9
So the maximum √ value of T on the disk x + y ≤ 1 is 4 , which occurs at
the two points ± 23 , − 12 on the boundary, and the minimum value of T on
the disk is − 14 which occurs at the critical point 0, 21 in the interior of R. 

From this example, we see that we use the following procedure for deter-
mining the absolute maximum and absolute minimum of a function on a closed
and bounded domain.

• Find all critical points of the function in the interior of the domain.
• Find all the critical points of the function on the boundary of the domain.
Working on the boundary of the domain reduces this part of the problem
to one or more single variable optimization problems. Note that there
may be endpoints on portions of the boundary that need to be considered.

• Evaluate the function at each of the points found in Steps 1 and 2.

• The maximum value of the function is the largest value obtained in Step
3, and the minimum value of the function is the smallest value obtained
in Step 3.
Activity 10.7.6 Let f (x, y) = x2 − 3y 2 − 4x + 6y with triangular domain R
whose vertices are at (0, 0), (4, 0), and (0, 4). The domain R and a graph of f
on the domain appear in Figure 10.7.10.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 181

z
y 5
4

1 2 2 1 x 0
y 3 3
3 -5 4

-10
2
-15

1
-20

x
1 2 3 4

Figure 10.7.10 The domain of f (x, y) = x2 − 3y 2 − 4x + 6y and its graph.

a. Find all of the critical points of f in R.

b. Parameterize the horizontal leg of the triangular domain, and find the
critical points of f on that leg. (Hint: You may need to consider end-
points.)
c. Parameterize the vertical leg of the triangular domain, and find the crit-
ical points of f on that leg. (Hint: You may need to consider endpoints.)

d. Parameterize the hypotenuse of the triangular domain, and find the crit-
ical points of f on the hypotenuse. (Hint: You may need to consider
endpoints.)
e. Find the absolute maximum and absolute minimum values of f on R.

10.7.4 Summary
• To find the extrema of a function f = f (x, y), we first find the critical
points, which are points where one of the partials of f fails to exist, or
where fx = 0 and fy = 0.

• The Second Derivative Test helps determine whether a critical point is a

local maximum, local minimum, or saddle point.
• If f is defined on a closed and bounded domain, we find the absolute
maxima and minima by finding the critical points in the interior of the
domain, finding the critical points on the boundary, and testing the value
of f at both sets of critical points.

10.7.5 Exercises
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 182

1. The function 2
k(x, y) = e−y cos(5x)
has a critical point at (0, 0).
What is the value of D at this critical point? D =
What type of critical point is it? ( maximum  minimum  sad-
dle point  point with unknown behavior)
Answer 1. 2 · 25
Answer 2. maximum
2. Consider the function

f (x, y) = (18x − x2 )(18y − y 2 ).

Find and classify all critical points of the function. If there are more
blanks than critical points, leave the remaining entries blank.
fx =
fy =
fxx =
fxy =
fyy =
There are several critical points to be listed. List them lexicograhically,
that is in ascending order by x-coordinates, and for equal x-coordinates
in ascending order by y-coordinates (e.g., (1,1), (2, -1), (2, 3) is a correct
order)
The critical point with the smallest x-coordinate is
( , ) Classification:
(local minimum, local maximum, saddle point, cannot be determined)
The critical point with the next smallest x-coordinate is
( , ) Classification:
(local minimum, local maximum, saddle point, cannot be determined)
The critical point with the next smallest x-coordinate is
( , ) Classification:
(local minimum, local maximum, saddle point, cannot be determined)
The critical point with the next smallest x-coordinate is
( , ) Classification:
(local minimum, local maximum, saddle point, cannot be determined)
The critical point with the next smallest x-coordinate is
( , ) Classification:
(local minimum, local maximum, saddle point, cannot be determined)


Answer 1. (18 − 2x) 18y − y 2


Answer 2. 18x − x2 (18 − 2y)


Answer 3. −2 18y − y 2
Answer 4. (18 − 2x)(18 − 2y)


Answer 5. −2 18x − x2
Answer 6. 0
Answer 7. 0
Answer 8. saddle point
Answer 9. 0
Answer 10. 18
Answer 11. saddle point
Answer 12. 9
Answer 13. 9
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 183

Answer 14. local maximum

Answer 15. 18
Answer 16. 0
Answer 17. saddle point
Answer 18. 18
Answer 19. 18
Answer 20. saddle point
3. Suppose f (x, y) = xy − ax − by.
(A) How many local minimum points does f have in R2 ? (The answer
is an integer).
(B) How many local maximum points does f have in R2 ?
(C) How many saddle points does f have in R2 ?
Answer 1. 0
Answer 2. 0
Answer 3. 1
4. Let f (x, y) = 1/x + 2/y + 3x + 4y in the region R where x, y > 0.
Explain why f must have a global minimum at some point in R (note
that R is unbounded---how does this influence your explanation?). Then
find the global minimum.
minimum =
√ √
Answer. 2 3 · 1 + 2 4 · 2
5. Each of the following functions has at most one critical point. Graph a few
level curves and a few gradients and, on this basis alone, decide whether
the critical point is a local maximum, a local minimum, a saddle point,
or that there is no critical point.
2 2
For f (x, y) = e−2x −4y , type of critical point: ( Local Maximum
 Local Minimum  Saddle Point  No Critical Point)
2 2
For f (x, y) = e2x −4y , type of critical point: ( Local Maximum
 Local Minimum  Saddle Point  No Critical Point)
For f (x, y) = 2x2 +4y 2 +4, type of critical point: ( Local Maximum
 Local Minimum  Saddle Point  No Critical Point)
For f (x, y) = 2x2 +4y +4, type of critical point: ( Local Maximum
 Local Minimum  Saddle Point  No Critical Point)
Answer 1. Local Maximum
Answer 2. Saddle Point
Answer 3. Local Minimum
Answer 4. No Critical Point
6. Find the absolute minimum and absolute maximum of

f (x, y) = 9 − 6x + 8y

on the closed triangular region with vertices (0, 0), (8, 0) and (8, 12).
List the minimum/maximum values as well as the point(s) at which
they occur. If a min or max occurs at multiple points separate the points
with commas.
Minimum value:
Occurs at
Maximum value:
Occurs at
Answer 1. −39
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 184

Answer 2. (8, 0)
Answer 3. 57
Answer 4. (8, 12)
7. Find the maximum and minimum values of f (x, y) = xy on the ellipse
5x2 + y 2 = 3.
maximum value =
minimum value =
Answer 1. 0.670820393249937
Answer 2. −0.670820393249937
8. Find A and B so that f (x, y) = x2 + Ax + y 2 + B has a local minimum
at the point (7, 0), with z-coordinate 5.
A=
B=
Answer 1. −14
Answer 2. 5 + 7 · 7
9. The contours of a function f are shown in the figure below.

For each of the points shown, indicate whether you think it is a local
maximum, local minimum, saddle point, or none of these.
(a) Point P is ( a local maximum  a local minimum  a
saddle point  none of these)
(b) Point Q is ( a local maximum  a local minimum  a
saddle point  none of these)
(c) Point R is ( a local maximum  a local minimum  a
saddle point  none of these)
(d) Point S is ( a local maximum  a local minimum  a saddle
point  none of these)
Answer 1. a local maximum
Answer 2. a local maximum
Answer 3. none of these
Answer 4. a saddle point
10. Consider the three points (5, 1), (7, 3), and (9, 4).
(a) Supposed that at (5, 1), we know that fx = fy = 0 and fxx = 0,
fyy = 0, and fxy > 0. What can we conclude about the behavior of this
function near the point (5, 1)? ( (5,1) is a local maximum  (5,1) is
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 185

a local minimum  (5,1) is a saddle point  (5,1) is a none of these)

(b) Supposed that at (7, 3), we know that fx = fy = 0 and fxx > 0,
fyy < 0, and fxy = 0. What can we conclude about the behavior of this
function near the point (7, 3)? ( (7,3) is a local maximum  (7,3) is
a local minimum  (7,3) is a saddle point  (7,3) is a none of these)
(c) Supposed that at (9, 4), we know that fx = fy = 0 and fxx < 0,
fyy = 0, and fxy > 0. What can we conclude about the behavior of this
function near the point (9, 4)? ( (9,4) is a local maximum  (9,4) is
a local minimum  (9,4) is a saddle point  (9,4) is a none of these)
Using this information, on a separate sheet of paper sketch a possible
contour diagram for f .
Answer 1. (5,1) is a saddle point
Answer 2. (7,3) is a saddle point
Answer 3. (9,4) is a saddle point
11. Find three positive real numbers whose sum is 94 and whose product is a
maximum.
Enter the three numbers separated by commas:
Answer. 31.3333, 31.3333, 31.3333
12. A closed rectangular box has volume 22 cm3 . What are the lengths of the
edges giving the minimum surface area?
lengths =
(Give the three lengths as a comma separated list.)
Answer. 2.80204, 2.80204, 2.80204
13. An open rectangular box has volume 26 cm3 . What are the lengths of the
edges giving the minimum surface area?
lengths =
(Give the three lengths as a comma separated list.)
Answer. 3.73251, 3.73251, 1.86626
14. What is the shortest distance from the surface xy + 6x + z 2 = 41 to the
origin?
distance =
√
Answer. 29
15. Find the volume of the largest rectangular box with edges parallel to the
axes that can be inscribed in the ellipsoid

x2 y2 z2
+ + =1
16 81 4
Hint: By symmetry, you can restrict your attention to the first octant
(where x, y, z ≥ 0), and assume your volume has the form V = 8xyz.
Then arguing by symmetry, you need only look for points which achieve
the maximum which lie in the first octant.
Maximum volume:
Answer. 110.851251684408
16. Design a rectangular milk carton box of width w, length l, and height h
which holds 500 cm3 of milk. The sides of the box cost 1 cent/cm2 and
the top and bottom cost 2 cent/cm2 . Find the dimensions of the box that
minimize the total cost of materials used.
dimensions =
(Enter your answer as a comma separated list of lengths.)
Answer. 12.5992, 6.2996, 6.29962
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 186

17. Respond to each of the following prompts to solve the given optimization
problem.
a. Let f (x, y) = sin(x) + cos(y). Determine the absolute maximum
and minimum values of f . At what points do these extreme values
occur?
b. For a certain differentiable function F of two variables x and y, its
partial derivatives are

Fx (x, y) = x2 − y − 4 and Fy (x, y) = −x + y − 2.

Find each of the critical points of F , and classify each as a local

maximum, local minimum, or a saddle point.
c. Determine all critical points of T (x, y) = 48 + 3xy − x2 y − xy 2 and
classify each as a local maximum, local minimum, or saddle point.
x2
d. Find and classify all critical points of g(x, y) = 2 + 3y 3 + 9y 2 −
3xy + 9y − 9x
2
−2y 2
e. Find and classify all critical points of z = f (x, y) = ye−x .

f. Determine the absolute maximum and absolute minimum of f (x, y) =

2 + 2x + 2y − x2 − y 2 on the triangular plate in the first quadrant
bounded by the lines x = 0, y = 0, and y = 9 − x.
g. Determine the absolute maximum and absolute minimum of f (x, y) =
2 + 2x + 2y − x2 − y 2 over the closed disk of points (x, y) such that
(x − 1)2 + (y − 1)2 ≤ 1.
h. Find the point on the plane z = 6 − 3x − 2y that lies closest to the
origin.
18. If a continuous function f of a single variable has two critical numbers c1
and c2 at which f has relative maximum values, then f must have another
critical number c3 , because ``it is impossible to have two mountains with-
out some sort of valley in between. The other critical point can be a saddle
point (a pass between the mountains) or a local minimum (a true valley)."
(From Calculus in Vector Spaces by Lawrence J. Corwin and Robert H.
Szczarb.) Consider the function f defined by f (x, y) = 4x2 ey − 2x4 − e4y .
(From Ira Rosenholz in the Problems Section of the Mathematics Maga-
zine, Vol. 60 NO. 1, February 1987.) Show that f has exactly two critical
points, and that f has relative maximum values at each of these critical
points. Explain how this function f illustrates that it really is possible to
have two mountains without some sort of valley in between. Use appro-
priate technology to draw the surface defined by f to see graphically how
this happens.
19. If a continuous function f of a single variable has exactly one critical
number with a relative maximum at that critical point, then the value
of f at that critical point is an absolute maximum. In this exercise we
see that the same is not always true for functions of two variables. Let
f (x, y) = 3xey −x3 −e3y (from “``The Only Critical Point in Town"” Test
by Ira Rosenholz and Lowell Smylie in the Mathematics Magazine, VOL
58 NO 3 May 1985.). Show that f has exactly one critical point, has a
relative maximum value at that critical point, but that f has no absolute
maximum value. Use appropriate technology to draw the surface defined
by f to see graphically how this happens.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 187

20. A manufacturer wants to procure rectangular boxes to ship its product.

The boxes must contain 20 cubic feet of space. To be durable enough to
ensure the safety of the product, the material for the sides of the boxes will
cost $0.10 per square foot, while the material for the top and bottom will
cost $0.25 per square foot. In this activity we will help the manufacturer
determine the box of minimal cost.
a. What quantities are constant in this problem? What are the vari-
ables in this problem? Provide appropriate variable labels. What,
if any, restrictions are there on the variables?
b. Using your variables from (a), determine a formula for the total cost
C of a box.
c. Your formula in part (b) might be in terms of three variables. If so,
find a relationship between the variables, and then use this relation-
ship to write C as a function of only two independent variables.
d. Find the dimensions that minimize the cost of a box. Be sure to
verify that you have a minimum cost.
21. A rectangular box with length x, width y, and height z is being built.
The box is positioned so that one corner is stationed at the origin and the
box lies in the first octant where x, y, and z are all positive. There is an
added constraint on how the box is constructed: it must fit underneath
the plane with equation x + 2y + 3z = 6. In fact, we will assume that the
corner of the box “opposite” the origin must actually lie on this plane.
The basic problem is to find the maximum volume of the box.
a. Sketch the plane x + 2y + 3z = 6, as well as a picture of a potential
box. Label everything appropriately.
b. Explain how you can use the fact that one corner of the box lies on
the plane to write the volume of the box as a function of x and y
only. Do so, and clearly show the formula you find for V (x, y).
c. Find all critical points of V . (Note that when finding the critical
points, it is essential that you factor first to make the algebra easier.)
d. Without considering the current applied nature of the function V ,
classify each critical point you found above as a local maximum,
local minimum, or saddle point of V .
e. Determine the maximum volume of the box, justifying your answer
completely with an appropriate discussion of the critical points of
the function.

f. Now suppose that we instead stipulated that, while the vertex of

the box opposite the origin still had to lie on the plane, we were
only going to permit the sides of the box, x and y, to have values
in a specified range (given below). That is, we now want to find the
maximum value of V on the closed, bounded region
1
≤ x ≤ 1, 1 ≤ y ≤ 2.
2
Find the maximum volume of the box under this condition, justify-
ing your answer fully.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 188

22. The airlines place restrictions on luggage that can be carried onto planes.
• A carry-on bag can weigh no more than 40 lbs.

• The length plus width plus height of a bag cannot exceed 45 inches.
• The bag must fit in an overhead bin.
Let x, y, and z be the length, width, and height (in inches) of a carry on
bag. In this problem we find the dimensions of the bag of largest volume,
V = xyz, that satisfies the second restriction. Assume that we use all
45 inches to get a maximum volume. (Note that this bag of maximum
volume might not satisfy the third restriction.)
a. Write the volume V = V (x, y) as a function of just the two variables
x and y.

b. Explain why the domain over which V is defined is the triangular

region R with vertices (0,0), (45,0), and (0,45).
c. Find the critical points, if any, of V in the interior of the region R.
d. Find the maximum value of V on the boundary of the region R,
and then determine the dimensions of a bag with maximum volume
on the entire region R. (Note that most carry-on bags sold today
measure 22 by 14 by 9 inches with a volume of 2772 cubic inches,
so that the bags will fit into the overhead bins.)
23.
According to The Song of Insects byTable 10.7.11 Crickets chirping.
G.W. Pierce (Harvard College Press,
1948) the sound of striped ground x y
crickets chirping, in number of chirps 20.0 88.6
per second, is related to the temper- 16.0 71.6
ature. So the number of chirps per 19.8 93.3
second could be a predictor of tem- 18.4 84.3
perature. The data Pierce collected 17.1 80.6
is shown in Table 10.7.11., where x 15.5 75.2
is the (average) number of chirps per 14.7 69.7
second and y is the temperature in
17.1 82.0
degrees Fahrenheit.
A scatterplot of the data would show 15.4 69.4
that, while the relationship between 16.2 83.3
x and y is not exactly linear, it looks 15.0 79.6
to have a linear pattern. It could 17.2 82.6
be that the relationship is really lin- 16.0 80.6
ear but experimental error causes the 17.0 83.5
data to be slightly inaccurate. Or 14.4 76.3
perhaps the data is not linear, but
only approximately linear.
If we want to use the data to make predications, then we need to fit
a curve of some kind to the data. Since the cricket data appears roughly
linear, we will fit a linear function f of the form f (x) = mx+b to the data.
We will do this in such a way that we minimize the sums of the squares
of the distances between the y values of the data and the corresponding
y values of the line defined by f . This type of fit is called a least squares
approximation. If the data is represented by the points (x1 , y1 ), (x2 , y2 ),
. . ., (xn , yn ), then the square of the distance between yi and f (xi ) is
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 189

(f (xi ) − yi )2 = (mxi + b − yi )2 . So our goal is to minimize the sum of

these squares, of minimize the function S defined by
n
X
S(m, b) = (mxi + b − yi )2 .
i=1

a. Calculate Sm and Sb .
b. Solve the system Sm (m, b) = 0 and Sb (m, b) = 0 to show that the
critical point satisfies
Pn Pn Pn
n ( i=1 xi yi ) − ( i=1 xi ) ( i=1 yi )
m= Pn Pn 2
n ( i=1 x2i ) − ( i=1 xi )
Pn Pn 2

P n Pn
( i=1 yi ) i=1 xi − ( i=1 xi ) ( i=1 xi yi )
b= Pn Pn 2 .
2
n ( i=1 xi ) − ( i=1 xi )

(Hint: Don’t be daunted by these expressions, the system Sm (m, b) =

0 and Sb (m, b) = 0 is a system of two linearPequations inP the un-
n 2 n
knowns
Pn m and b. It might
Pnbe easier to let r = x
i=1 i , s = i=1 xi ,
t = y
i=1 i , and u = x y
i=1 i i and write your equations using
these constants.)
c. Use the Second Derivative Test to explain why the critical point
gives a local minimum. Can you then explain why the critical point
gives an absolute minimum?
d. Use the formula from part (b) to find the values of m and b that
give the line of best fit in the least squares sense to the cricket data.
Draw your line on the scatter plot to convince yourself that you have
a well-fitting line.

10.8 Constrained Optimization: Lagrange Mul-

tipliers
Motivating Questions
• What geometric condition enables us to optimize a function f = f (x, y)
subject to a constraint given by g(x, y) = k, where k is a constant?

• How can we exploit this geometric condition to find the extreme values
of a function subject to a constraint?
We previously considered how to find the extreme values of functions on
both unrestricted domains and on closed, bounded domains. Other types of
optimization problems involve maximizing or minimizing a quantity subject
to an external constraint. In these cases the extreme values frequently won’t
occur at the points where the gradient is zero, but rather at other points that
satisfy an important geometric condition. These problems are often called con-
strained optimization problems and can be solved with the method of Lagrange
Multipliers, which we study in this section.
Preview Activity 10.8.1 According to U.S. postal regulations, the girth
plus the length of a parcel sent by mail may not exceed 108 inches, where
by “girth” we mean the perimeter of the smallest end. Our goal is to find
the largest possible volume of a rectangular parcel with a square end that
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 190

can be sent by mail. (We solved this applied optimization problem in single
variable Active Calculus, so it may look familiar. We take a different approach
in this section, and this approach allows us to view most applied optimization
problems from single variable calculus as constrained optimization problems,
as well as provide us tools to solve a greater variety of optimization problems.)
If we let x be the length of the side of one square end of the package and y
the length of the package, then we want to maximize the volume f (x, y) = x2 y
of the box subject to the constraint that the girth (4x) plus the length (y) is
as large as possible, or 4x + y = 108. The equation 4x + y = 108 is thus an
external constraint on the variables.
a. The constraint equation involves the function g that is given by

g(x, y) = 4x + y.

Explain why the constraint is a contour of g, and is therefore a two-

dimensional curve.
120 y

100

80
A
60
C
40
5000
20000
20 D
1000 10000
B x
10 20 30 40

Figure 10.8.1 Contours of f and the constraint equation g(x, y) = 108.

b. Figure 10.8.1 shows the graph of the constraint equation g(x, y) = 108
along with a few contours of the volume function f . Since our goal is to
find the maximum value of f subject to the constraint g(x, y) = 108, we
want to find the point on our constraint curve that intersects the contours
of f at which f has its largest value.
i. Points A and B in Figure 10.8.1 lie on a contour of f and on the
constraint equation g(x, y) = 108. Explain why neither A nor B
provides a maximum value of f that satisfies the constraint.
ii. Points C and D in Figure 10.8.1 lie on a contour of f and on the
constraint equation g(x, y) = 108. Explain why neither C nor D
provides a maximum value of f that satisfies the constraint.
iii. Based on your responses to parts i. and ii., draw the contour of f
on which you believe f will achieve a maximum value subject to the
constraint g(x, y) = 108. Explain why you drew the contour you
did.

c. Recall that g(x, y) = 108 is a contour of the function g, and that the
gradient of a function is always orthogonal to its contours. With this in
mind, how should ∇f and ∇g be related at the optimal point? Explain.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 191

10.8.1 Constrained Optimization and Lagrange Multipli-

ers
In Preview Activity 10.8.1, we considered an optimization problem where there
is an external constraint on the variables, namely that the girth plus the length
of the package cannot exceed 108 inches. We saw that we can create a function
g from the constraint, specifically g(x, y) = 4x + y. The constraint equation is
then just a contour of g, g(x, y) = c, where c is a constant (in our case 108).
Figure 10.8.2 illustrates that the volume function f is maximized, subject to
the constraint g(x, y) = c, when the graph of g(x, y) = c is tangent to a contour
of f . Moreover, the value of f on this contour is the sought maximum value.
120 y

100

80

60

40

20
x
10 20 30 40
Figure 10.8.2 Contours of f and the constraint contour.
To find this point where the graph of the constraint is tangent to a contour
of f , recall that ∇f is perpendicular to the contours of f and ∇g is perpen-
dicular to the contour of g. At such a point, the vectors ∇g and ∇f are
parallel, and thus we need to determine the points where this occurs. Recall
that two vectors are parallel if one is a nonzero scalar multiple of the other, so
we therefore look for values of a parameter λ that make

∇f = λ∇g. (10.8.1)

The constant λ is called a Lagrange multiplier.

To find the values of λ that satisfy (10.8.1) for the volume function in
Preview Activity 10.8.1, we calculate both ∇f and ∇g. Observe that

∇f = 2xyi + x2 j and ∇g = 4i + j,

and thus we need a value of λ so that

2xyi + x2 j = λ(4i + j).

Equating components in the most recent equation and incorporating the

original constraint, we have three equations

2xy = λ(4) (10.8.2)

2
x = λ(1) (10.8.3)
4x + y = 108 (10.8.4)

in the three unknowns x, y, and λ. First, note that if λ = 0, then equation

(10.8.3) shows that x = 0. From this, Equation (10.8.4) tells us that y = 108.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 192

So the point (0, 108) is a point we need to consider. Next, provided that λ 6= 0
(from which it follows that x 6= 0 by Equation (10.8.3)), we may divide both
sides of Equation (10.8.2) by the corresponding sides of (10.8.3) to eliminate
λ, and thus find that
2y
= 4, so
x
y = 2x.

Substituting into Equation (10.8.4) gives us

4x + 2x = 108

or
x = 18.
Thus we have y = 2x = 36 and λ = x2 = 324 as another point to consider.
So the points at which the gradients of f and g are parallel, and thus at which
f may have a maximum or minimum subject to the constraint, are (0, 108) and
(18, 36). By evaluating the function f at these points, we see that we maximize
the volume when the length of the square end of the box is 18 inches and the
length is 36 inches, for a maximum volume of f (18, 36) = 11664 cubic inches.
Since f (0, 108) = 0, we obtain a minimum value at this point.
We summarize the process of Lagrange multipliers as follows.

The method of Lagrange multipliers.

The general technique for optimizing a function f = f (x, y) subject to a

constraint g(x, y) = c is to solve the system ∇f = λ∇g and g(x, y) = c
for x, y, and λ. We then evaluate the function f at each point (x, y)
that results from a solution to the system in order to find the optimum
values of f subject to the constraint.

Activity 10.8.2 A cylindrical soda can holds about 355 cc of liquid. In this
activity, we want to find the dimensions of such a can that will minimize the
surface area. For the sake of simplicity, assume the can is a perfect cylinder.
a. What are the variables in this problem? Based on the context, what
restriction(s), if any, are there on these variables?
b. What quantity do we want to optimize in this problem? What equation
describes the constraint? (You need to decide which of these functions
plays the role of f and which plays the role of g in our discussion of
Lagrange multipliers.)

c. Find λ and the values of your variables that satisfy Equation (10.8.1) in
the context of this problem.
d. Determine the dimensions of the pop can that give the desired solution
to this constrained optimization problem.
The method of Lagrange multipliers also works for functions of more than
two variables.
Activity 10.8.3 Use the method of Lagrange multipliers to find the dimensions
of the least expensive packing crate with a volume of 240 cubic feet when the
material for the top costs $2 per square foot, the bottom is $3 per square foot
and the sides are $1.50 per square foot.
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 193

The method of Lagrange multipliers also works for functions of three vari-
ables. That is, if we have a function f = f (x, y, z) that we want to optimize
subject to a constraint g(x, y, z) = k, the optimal point (x, y, z) lies on the
level surface S defined by the constraint g(x, y, z) = k. As we did in Preview
Activity 10.8.1, we can argue that the optimal value occurs at the level surface
f (x, y, z) = c that is tangent to S. Thus, the gradients of f and g are parallel
at this optimal point. So, just as in the two variable case, we can optimize
f = f (x, y, z) subject to the constraint g(x, y, z) = k by finding all points
(x, y, z) that satisfy ∇f = λ∇g and g(x, y, z) = k.

10.8.2 Summary
• The extrema of a function f = f (x, y) subject to a constraint g(x, y) = c
occur at points for which the contour of f is tangent to the curve that
represents the constraint equation. This occurs when

∇f = λ∇g.

• We use the condition ∇f = λ∇g to generate a system of equations,

together with the constraint g(x, y) = c, that may be solved for x, y, and
λ. Once we have all the solutions, we evaluate f at each of the (x, y)
points to determine the extrema.

10.8.3 Exercises
1. Use Lagrange multipliers to find the maximum and minimum values of
f (x, y) = 3x − 4y subject to the constraint x2 + 3y 2 = 129, if such values
exist.
maximum =
minimum =
(For either value, enter DNE if there is no such value.)
Answer 1. 3 · 3 · 3 + 4 · 4
Answer 2. −3 · 3 · 3 − 4 · 4
2. Use Lagrange multipliers to find the maximum and minimum values of
f (x, y) = x2 y+3y 2 −y, subject to the constraint x2 +y 2 ≤ 38.3333333333333
maximum =
minimum =
(For either value, enter DNE if there is no such value.)
Answer 1. 137.926
Answer 2. −59.2593
3. Find the absolute maximum and minimum of the function f (x, y) = x2 +
y 2 subject to the constraint x4 + y 4 = 6561.
As usual, ignore unneeded answer blanks, and list points in lexico-
graphic order.
Absolute minimum value:
attained at ( , ), ( ,
),
( , ), ( , ).
Absolute maximum value:
attained at ( , ), ( ,
),
( , ), ( , ).
Answer 1. 81
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 194

Answer 2. −9
Answer 3. 0
Answer 4. 0
Answer 5. −9
Answer 6. 0
Answer 7. 9
Answer 8. 9
Answer 9. 0
Answer 10. 114.551298552221
Answer 11. −7.56806773728343
Answer 12. −7.56806773728343
Answer 13. −7.56806773728343
Answer 14. 7.56806773728343
Answer 15. 7.56806773728343
Answer 16. −7.56806773728343
Answer 17. 7.56806773728343
Answer 18. 7.56806773728343
4. Find the absolute maximum and minimum of the function f (x, y) = x2 −
y 2 subject to the constraint x2 + y 2 = 361.
As usual, ignore unneeded answer blanks, and list points in lexico-
graphic order.
Absolute minimum value:
attained at ( , ) and ( ,
).
Absolute maximum value:
attained at ( , ) and ( ,
).
Answer 1. −361
Answer 2. 0
Answer 3. −19
Answer 4. 0
Answer 5. 19
Answer 6. 361
Answer 7. −19
Answer 8. 0
Answer 9. 19
Answer 10. 0
5. Find the minimum distance from the point (1, 1, 11) to the paraboloid
given by the equation z = x2 + y 2 .
Minimum distance =
Note: If you need to find roots of a polynomial of degree ≥ 3, you
may want to use a calculator of computer to do so numerically. Also be
sure that you can give a geometric justification for your answer.
√
Answer. 3.53771
6. For each value of λ the function h(x, y) = x2 + y 2 − λ(2x + 8y − 20) has
a minimum value m(λ).
(a) Find m(λ)
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 195

m(λ) =
(Use the letter L for λ in your expression.)
(b) For which value of λ is m(λ) the largest, and what is that maximum
value?
λ=
maximum m(λ) =
(c) Find the minimum value of f (x, y) = x2 + y 2 subject to the con-
straint 2x+8y = 20 using the method of Lagrange multipliers and evaluate
λ.
minimum f =
λ=
(How are these results related to your result in part (b)?)
(2·2+8·8)L2
Answer 1. 20L − 4
2·20
Answer 2. 2·2+8·8
20·20
Answer 3. 2·2+8·8
20·20
Answer 4. 2·2+8·8
2·20
Answer 5. 2·2+8·8

7. The plane x + y + 2z = 6 intersects the paraboloid z = x2 + y 2 in an

ellipse. Find the points on this ellipse that are nearest to and farthest
from the origin.
Point farthest away occurs at
( , ,
).
Point nearest occurs at
( , ,
).
Answer 1. −1.5
Answer 2. −1.5
Answer 3. 4.5
Answer 4. 1
Answer 5. 1
Answer 6. 2
8. Find the maximum and minimum values of the function f (x, y, z) =
x2 y 2 z 2 subject to the constraint x2 + y 2 + z 2 = 64.
Maximum value is , occuring at
points (positive integer or "infinitely many").
Minimum value is , occuring at
points (positive integer or "infinitely many").
Answer 1. 9709.03703703704
Answer 2. 8
Answer 3. 0
Answer 4. infinitely many
9. Find the maximum and minimum values of the function f (x, y, z, t) =
x + y + z + t subject to the constraint x2 + y 2 + z 2 + t2 = 100.
Maximum value is , occuring at
points (positive integer or "infinitely many").
Minimum value is , occuring at
points (positive integer or "infinitely many").
Answer 1. 20
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 196

Answer 2. 1
Answer 3. −20
Answer 4. 1
10. Find the maximum and minimum volumes of a rectangular box whose
surface area equals 7000 square cm and whose edge length (sum of lengths
of all edges) is 440 cm.
Hint: It can be deduced that the box is not a cube, so if x, y, and z
are the lengths of the sides, you may want to let x represent a side with
x 6= y and x 6= z.
Maximum value is ,
occuring at ( ,
, ).
Minimum value is ,
occuring at ( ,
, ).
Answer 1. 34481.4814814815
Answer 2. 63.3333333333333
Answer 3. 23.3333333333333
Answer 4. 23.3333333333333
Answer 5. 25000
Answer 6. 10
Answer 7. 50
Answer 8. 50
P3 P3
11. (a) If i=1 xi = 4, find the values of x1 , x2 , x3 making i=1 xi 2 mini-
mum.
x1 , x2 , x3 =
(Give your values as a comma separated list.)
Pn(b) Generalize
2
the
Pnresult of part (a) to find the minimum value of
i=1 x i subject to i=1 xi = 4.
minimum value =
Answer 1. 1.33333, 1.33333, 1.33333
4·4
Answer 2. n
12. The Cobb-Douglas production function is used in economics to model
production levels based on labor and equipment. Suppose we have a
specific Cobb-Douglas function of the form

f (x, y) = 50x0.4 y 0.6 ,

where x is the dollar amount spent on labor and y the dollar amount spent
on equipment. Use the method of Lagrange multipliers to determine how
much should be spent on labor and how much on equipment to maximize
productivity if we have a total of 1.5 million dollars to invest in labor and
equipment.
13. Use the method of Lagrange multipliers to find the point on the line
x − 2y = 5 that is closest to the point (1, 3). To do so, respond to the
following prompts.
a. Write the function f = f (x, y) that measures the square of the
distance from (x, y) to (1, 3). (The extrema of this function are the
same as the extrema of the distance function, but f (x, y) is simpler
to work with.)
CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 197

b. What is the constraint g(x, y) = c?

c. Write the equations resulting from ∇f = λ∇g and the constraint.
Find all the points (x, y) satisfying these equations.
d. Test all the points you found to determine the extrema.
14. Apply the Method of Lagrange Multipliers solve each of the following
constrained optimization problems.
a. Determine the absolute maximum and absolute minimum values of
f (x, y) = (x − 1)2 + (y − 2)2 subject to the constraint that x2 + y 2 =
16.
b. Determine the points on the sphere x2 +y 2 +z 2 = 4 that are closest to
and farthest from the point (3, 1, −1). (As in the preceding exercise,
you may find it simpler to work with the square of the distance
formula, rather than the distance formula itself.)
c. Find the absolute maximum and minimum of f (x, y, z) = x2 +y 2 +z 2
subject to the constraint that (x−3)2 +(y+2)2 +(z−5)2 ≤ 16. (Hint:
here the constraint is a closed, bounded region. Use the boundary
of that region for applying Lagrange Multipliers, but don’t forget to
also test any critical values of the function that lie in the interior of
the region.)
15. In this exercise we consider how to apply the Method of Lagrange Mul-
tipliers to optimize functions of three variable subject to two constraints.
Suppose we want to optimize f = f (x, y, z) subject to the constraints
g(x, y, z) = c and h(x, y, z) = k. Also suppose that the two level surfaces
g(x, y, z) = c and h(x, y, z) = k intersect at a curve C. The optimum
point P = (x0 , y0 , z0 ) will then lie on C.
a. Assume that C can be represented parametrically by a vector-valued
−−→
function r = r(t). Let OP = r(t0 ). Use the Chain Rule applied to
f (r(t)), g(r(t)), and h(r(t)), to explain why

∇f (x0 , y0 , z0 ) · r0 (t0 ) = 0,
∇g(x0 , y0 , z0 ) · r0 (t0 ) = 0, and
∇h(x0 , y0 , z0 ) · r0 (t0 ) = 0.

Explain how this shows that ∇f (x0 , y0 , z0 ), ∇g(x0 , y0 , z0 ), and ∇h(x0 , y0 , z0 )

are all orthogonal to C at P . This shows that ∇f (x0 , y0 , z0 ), ∇g(x0 , y0 , z0 ),
and ∇h(x0 , y0 , z0 ) all lie in the same plane.

b. Assuming that ∇g(x0 , y0 , z0 ) and ∇h(x0 , y0 , z0 ) are nonzero and

not parallel, explain why every point in the plane determined by
∇g(x0 , y0 , z0 ) and ∇h(x0 , y0 , z0 ) has the form s∇g(x0 , y0 , z0 )+t∇h(x0 , y0 , z0 )
for some scalars s and t.
c. Parts (a.) and (b.) show that there must exist scalars λ and µ such
that

∇f (x0 , y0 , z0 ) = λ∇g(x0 , y0 , z0 ) + µ∇h(x0 , y0 , z0 ).

So to optimize f = f (x, y, z) subject to the constraints g(x, y, z) = c

and h(x, y, z) = k we must solve the system of equations

∇f (x, y, z) = λ∇g(x, y, z) + µ∇h(x, y, z),

CHAPTER 10. DERIVATIVES OF MULTIVARIABLE FUNCTIONS 198

g(x, y, z) = c, and
h(x, y, z) = k.

for x, y, z, λ, and µ.
Use this idea to find the maximum and minium values of f (x, y, z) =
x + 2y subject to the constraints y 2 + z 2 = 8 and x + y + z = 10.
16. There is a useful interpretation of the Lagrange multiplier λ. Assume
that we want to optimize a function f with constraint g(x, y) = c. Recall
that an optimal solution occurs at a point (x0 , y0 ) where ∇f = λ∇g. As
the constraint changes, so does the point at which the optimal solution
occurs. So we can think of the optimal point as a function of the parameter
c, that is x0 = x0 (c) and y0 = y0 (c). The optimal value of f subject
to the constraint can then be considered as a function of c defined by
f (x0 (c), y0 (c)). The Chain Rule shows that

df ∂f dx0 ∂f dy0
= + .
dc ∂x0 dc ∂y0 dc

a. Use the fact that ∇f = λ∇g at (x0 , y0 ) to explain why

df dg
=λ .
dc dc

b. Use the fact that g(x, y) = c to show that

df
= λ.
dc
Conclude that λ tells us the rate of change of the function f as the
parameter c increases (or by approximately how much the optimal
value of the function f will change if we increase the value of c by 1
unit).

c. Suppose that λ = 324 at the point where the package described

in Preview Activity 10.8.1 has its maximum volume. Explain in
context what the value 324 tells us about the package.
d. Suppose that the maximum value of a function f = f (x, y) subject
to a constraint g(x, y) = 100 is 236. When using the method of
Lagrange multipliers and solving ∇f = λ∇g, we obtain a value of
λ = 15 at this maximum. Find an approximation to the maximum
value of f subject to the constraint g(x, y) = 98.