From the free course College Algebra on Lykke

6.7 Exponential and Logarithmic Models

Key concepts: Exponential growth and decay models · Using exponential models to predict populations · Determining the domain of an exponential model · The mathematical constant e · Continuous growth and continuous decay · Continuous compounding in finance · Interpreting positive and negative continuous rates · Evaluating exponential models at a specified time · Comparing exponential population models · Modeling radioactive decay

By the end of this section the student can: Model exponential growth and decay; Use Newton’s Law of Cooling; Use logistic-growth models; Choose an appropriate model for data; Express an exponential model in base

6.7 Exponential and Logarithmic Models

A population that grows by the same percentage each year does not add the same number of individuals each year: as the population becomes larger, the number added becomes larger too. That feedback produces an exponential model, while temperature change, limited resources, and changing rates often require more specialized models.

An exponential model describes a quantity whose change is proportional to its current value. Exponential models are useful for populations, investments, biological samples, radioactive substances, temperature, and many other quantities. The central question is not merely whether a quantity increases or decreases, but how its rate of change behaves over time.

Modeling Exponential Growth and Decay

For rapid growth, an exponential growth function has the form

$$ A(t)=ae^{kt}, $$

where $a$ is the initial value, $e$ is Euler’s number, approximately $2.71828$, $k$ is a positive constant that determines the growth rate, and $t$ is time.

If $k>0$, then $e^{kt}$ increases as $t$ increases, so the model represents exponential growth. If the model is written as

$$ A(t)=ae^{-kt}, $$

with $k>0$, then the exponent is negative and the model represents exponential decay.

A related form uses a constant base:

$$ A(t)=ab^t. $$

Here $a$ is the initial value and $b$ is the growth factor. If $b>1$, the model represents growth; if $0<b<1$, it represents decay. The growth or decay rate is obtained from $b$:

$$ \text{growth rate}=b-1, $$

and

$$ \text{decay rate}=1-b. $$

For example, $b=1.06$ means a $6%$ increase per time unit, while $b=0.97$ means a $3%$ decrease per time unit.

Definition. Exponential growth occurs when a quantity increases by a constant percentage over equal time intervals. Exponential decay occurs when a quantity decreases by a constant percentage over equal time intervals.

A useful feature of an exponential model is its doubling time, the amount of time required for a growing quantity to double. For

$$ A(t)=ae^{kt}, $$

set $A(t)=2a$:

$$ 2a=ae^{kt}. $$

Divide by $a$:

$$ 2=e^{kt}. $$

Take the natural logarithm:

$$ \ln 2=kt. $$

Solve for $t$:

$$ t=\frac{\ln 2}{k}. $$

Example: Modeling population growth. A wildlife population begins with $500$ animals and grows continuously at a rate of $4%$ per year. Find a model and predict the population after $10$ years.

The initial value is

$$ a=500. $$

The continuous growth rate is

$$ k=0.04. $$

Substitute these values into $A(t)=ae^{kt}$:

$$ A(t)=500e^{0.04t}. $$

To find the population after $10$ years, substitute $t=10$:

$$ A(10)=500e^{0.04(10)}. $$

Multiply in the exponent:

$$ A(10)=500e^{0.4}. $$

Approximate the power of $e$:

$$ e^{0.4}\approx 1.49182. $$

Multiply:

$$ A(10)\approx 500(1.49182). $$

Therefore,

$$ A(10)\approx 745.91. $$

Since a population is counted in whole animals, the predicted population is approximately

$$ \boxed{746\text{ animals}}. $$

The domain of a population model is determined by the meaning of time. Algebraically, $A(t)=500e^{0.04t}$ is defined for every real number $t$, but a model describing population growth usually uses $t\geq 0$, because negative time represents dates before the model began. A model may also stop being realistic when resources, disease, migration, or environmental limits change the population behavior.

Using Newton’s Law of Cooling

Newton’s Law of Cooling states that the rate at which an object’s temperature changes is proportional to the difference between the object’s temperature and the surrounding temperature. The object cools rapidly when the temperature difference is large and slowly when the object approaches the surrounding temperature.

The model is

$$ T(t)=T_s+(T_0-T_s)e^{-kt}, $$

where $T(t)$ is the object’s temperature at time $t$, $T_s$ is the constant surrounding temperature, $T_0$ is the initial temperature, and $k>0$ is a constant determined by the situation.

The expression $T_0-T_s$ measures the object’s initial temperature difference from its surroundings. The factor $e^{-kt}$ reduces that difference over time. As $t$ becomes large, $e^{-kt}$ approaches $0$, so $T(t)$ approaches $T_s$.

Example: Finding the temperature of cooling coffee. A cup of coffee has an initial temperature of $180^\circ\text{F}$ in a room maintained at $70^\circ\text{F}$. After $20$ minutes, the coffee has cooled to $130^\circ\text{F}$. Find the temperature after $40$ minutes.

Start with Newton’s model:

$$ T(t)=T_s+(T_0-T_s)e^{-kt}. $$

Substitute $T_s=70$ and $T_0=180$:

$$ T(t)=70+(180-70)e^{-kt}. $$

Simplify:

$$ T(t)=70+110e^{-kt}. $$

Use the measurement $T(20)=130$:

$$ 130=70+110e^{-20k}. $$

Subtract $70$ from both sides:

$$ 60=110e^{-20k}. $$

Divide by $110$:

$$ \frac{60}{110}=e^{-20k}. $$

Reduce the fraction:

$$ \frac{6}{11}=e^{-20k}. $$

Take the natural logarithm of both sides:

$$ \ln\left(\frac{6}{11}\right)=\ln\left(e^{-20k}\right). $$

Use $\ln(e^x)=x$:

$$ \ln\left(\frac{6}{11}\right)=-20k. $$

Divide by $-20$:

$$ k=-\frac{1}{20}\ln\left(\frac{6}{11}\right). $$

Because $\ln(6/11)$ is negative, $k$ is positive, as required. Approximate:

$$ k\approx 0.03038. $$

Now find $T(40)$:

$$ T(40)=70+110e^{-0.03038(40)}. $$

Multiply:

$$ T(40)=70+110e^{-1.2152}. $$

Approximate the exponential:

$$ e^{-1.2152}\approx 0.2968. $$

Multiply:

$$ T(40)\approx 70+110(0.2968). $$

$$ T(40)\approx 70+32.648. $$

Therefore,

$$ \boxed{T(40)\approx 102.6^\circ\text{F}}. $$

A common mistake is to use $T(t)=T_0e^{-kt}$ for a cooling object. That model assumes the temperature approaches $0^\circ\text{F}$ or $0^\circ\text{C}$, not the room temperature. Newton’s model must include the surrounding temperature $T_s$.

Using Logistic Growth Models

Exponential growth cannot continue forever in a limited environment. A population may initially grow exponentially, but food, space, disease, and competition eventually slow the growth. A logistic growth model describes growth that begins rapidly and then levels off near a maximum sustainable population called the carrying capacity.

The logistic model is

$$ P(t)=\frac{c}{1+ae^{-bt}}, $$

where $c$ is the carrying capacity, $a$ is a constant determined by the initial population, and $b$ is a positive growth constant.

The graph of a logistic model is typically $S$-shaped. At first, the population grows slowly, then rapidly, and finally slowly again as it approaches $c$. The population does not generally equal $c$ at a finite time; instead,

$$ \lim_{t\to\infty}P(t)=c. $$

Example: Using a fish-farm logistic model. Suppose the fish population is modeled by

$$ P(t)=\frac{1000}{1+9e^{-0.6t}}, $$

where $t$ is measured in years. Find the initial population and the population after $2$ years.

To find the initial population, substitute $t=0$:

$$ P(0)=\frac{1000}{1+9e^{-0.6(0)}}. $$

$$ P(0)=\frac{1000}{1+9e^0}. $$

Since $e^0=1$:

$$ P(0)=\frac{1000}{1+9(1)}. $$

Simplify the denominator:

$$ P(0)=\frac{1000}{10}. $$

Therefore,

$$ \boxed{P(0)=100\text{ fish}}. $$

Now find the population after $2$ years:

$$ P(2)=\frac{1000}{1+9e^{-0.6(2)}}. $$

Multiply:

$$ P(2)=\frac{1000}{1+9e^{-1.2}}. $$

Approximate:

$$ e^{-1.2}\approx 0.3010. $$

Substitute:

$$ P(2)\approx\frac{1000}{1+9(0.3010)}. $$

Multiply:

$$ P(2)\approx\frac{1000}{1+2.709}. $$

Add:

$$ P(2)\approx\frac{1000}{3.709}. $$

Divide:

$$ P(2)\approx269.6. $$

Thus the predicted population is

$$ \boxed{270\text{ fish, approximately}}. $$

In this example, the carrying capacity is $c=1000$, not the initial population. The coefficient $9$ controls how the curve is positioned, while $0.6$ controls the growth behavior. Treating the model as exponential for all time would eventually predict populations larger than the environment can sustain.

Choosing an Appropriate Model for Data

A model should reflect the pattern in the data rather than merely pass through one or two points. Use an exponential model when the quantity changes by approximately the same percentage over equal time intervals. Use a linear model when it changes by approximately the same absolute amount. Use a logistic model when growth begins exponentially but later slows toward a limiting value.

A practical method is:

  1. Inspect the data and identify whether the changes are additive or multiplicative.
  2. For consecutive data values, compare differences and ratios.
  3. If the differences are roughly constant, consider a linear model.
  4. If the ratios are roughly constant, consider an exponential model.
  5. If the data rise and then level off near a maximum, consider a logistic model.
  6. Check the model against all available data and reject models that produce unreasonable long-term behavior.

Example: Choosing between linear and exponential models. Consider the data

$$ \begin{array}{c|ccccc} t & 0 & 1 & 2 & 3 & 4\ \hline A(t) & 100 & 120 & 144 & 173 & 207 \end{array} $$

The successive differences are

$$ 120-100=20, $$

$$ 144-120=24, $$

$$ 173-144=29, $$

$$ 207-173=34. $$

The differences are not constant. Now calculate the ratios:

$$ \frac{120}{100}=1.20, $$

$$ \frac{144}{120}=1.20, $$

$$ \frac{173}{144}\approx1.20, $$

$$ \frac{207}{173}\approx1.20. $$

The ratios are approximately constant, so an exponential model is appropriate. Use

$$ A(t)=ab^t. $$

Because $A(0)=100$,

$$ 100=ab^0. $$

Since $b^0=1$,

$$ 100=a. $$

The common ratio is approximately $b=1.20$, so

$$ \boxed{A(t)=100(1.20)^t}. $$

Check the model at $t=3$:

$$ A(3)=100(1.20)^3. $$

Evaluate the power:

$$ (1.20)^3=1.728. $$

Multiply:

$$ A(3)=100(1.728)=172.8. $$

This is close to the observed value $173$, confirming the model’s suitability.

Expressing an Exponential Model in Base $e$

Every positive-base exponential model can be expressed using base $e$. The key identity is

$$ b^t=e^{t\ln b}, $$

for $b>0$. Therefore,

$$ ab^t=ae^{t\ln b}. $$

Thus the model

$$ A(t)=ab^t $$

can be rewritten as

$$ A(t)=ae^{kt}, $$

where

$$ k=\ln b. $$

This conversion is useful because logarithms and continuous growth formulas naturally use base $e$. If $b>1$, then $\ln b>0$ and the converted model represents growth. If $0<b<1$, then $\ln b<0$ and the converted model represents decay.

Example: Converting a model to base $e$. Express

$$ A(t)=500(1.06)^t $$

in base $e$.

Use the identity

$$ b^t=e^{t\ln b}. $$

Here $b=1.06$, so

$$ (1.06)^t=e^{t\ln(1.06)}. $$

Substitute into the original model:

$$ A(t)=500e^{t\ln(1.06)}. $$

Approximate the natural logarithm:

$$ \ln(1.06)\approx0.05827. $$

Therefore,

$$ \boxed{A(t)\approx500e^{0.05827t}}. $$

The original model has a growth factor of $1.06$ per time unit, while the base-$e$ model has a continuous growth rate of approximately $0.05827$, or about $5.827%$ per time unit.

The continuous growth formula is

$$ A(t)=ae^{rt}, $$

where $a$ is the initial value, $r$ is the continuous growth rate per unit time, and $t$ is elapsed time. For financial applications, the same formula is written

$$ A(t)=Pe^{rt}, $$

where $P$ is the principal, or initial investment. If $r>0$, the account grows continuously; if $r<0$, it decays continuously.

Example: Continuous compounding. An account contains $P=$1200$ and earns a continuous annual interest rate of $5%$. Find its value after $3$ years.

Convert the percentage to a decimal:

$$ r=5%=0.05. $$

Use the continuous compounding formula:

$$ A(t)=Pe^{rt}. $$

Substitute $P=1200$, $r=0.05$, and $t=3$:

$$ A(3)=1200e^{0.05(3)}. $$

Multiply:

$$ A(3)=1200e^{0.15}. $$

Approximate:

$$ e^{0.15}\approx1.16183. $$

Multiply:

$$ A(3)\approx1200(1.16183). $$

Therefore,

$$ \boxed{A(3)\approx$1,394.20}. $$

The number $e$ is not an arbitrary calculator constant. It is the natural base for continuous growth because repeated compounding over increasingly short time intervals leads to an expression involving $e$. Continuous models therefore provide a unified way to describe population growth, decay, cooling, and continuously compounded finance.

Practice Problems

  1. A bacteria culture begins with $800$ cells and grows continuously at a rate of $3%$ per hour. Find a model and the population after $6$ hours.

  2. An object at $200^\circ\text{C}$ is placed in a room at $25^\circ\text{C}$. After $10$ minutes, its temperature is $125^\circ\text{C}$. Use Newton’s Law of Cooling to find its temperature after $20$ minutes.

  3. Convert the exponential model

$$ A(t)=750(0.92)^t $$

to base $e$.

Answers

  1. The model is

$$ A(t)=800e^{0.03t}. $$

After $6$ hours,

$$ A(6)=800e^{0.18}\approx957.6. $$

So the population is approximately

$$ \boxed{958\text{ cells}}. $$

  1. Start with

$$ T(t)=25+(200-25)e^{-kt} $$

and simplify:

$$ T(t)=25+175e^{-kt}. $$

Use $T(10)=125$:

$$ 125=25+175e^{-10k}, $$

$$ 100=175e^{-10k}, $$

$$ \frac{4}{7}=e^{-10k}, $$

$$ \ln\left(\frac{4}{7}\right)=-10k. $$

Thus

$$ k=-\frac{1}{10}\ln\left(\frac{4}{7}\right)\approx0.05596. $$

Then

$$ T(20)=25+175e^{-0.05596(20)} $$

and

$$ T(20)\approx25+175e^{-1.1192}\approx25+57.2. $$

Therefore,

$$ \boxed{T(20)\approx82.2^\circ\text{C}}. $$

  1. Use

$$ b^t=e^{t\ln b}. $$

Then

$$ A(t)=750(0.92)^t $$

becomes

$$ A(t)=750e^{t\ln(0.92)}. $$

Since

$$ \ln(0.92)\approx-0.08338, $$

the base-$e$ form is

$$ \boxed{A(t)\approx750e^{-0.08338t}}. $$

Content adapted from OpenStax College Algebra 2e by Jay Abramson et al., licensed CC BY 4.0.

6.7 Exponential and Logarithmic Models - College Algebra - image 1
6.7 Exponential and Logarithmic Models - College Algebra - image 1
6.7 Exponential and Logarithmic Models - College Algebra - image 2
6.7 Exponential and Logarithmic Models - College Algebra - image 2
6.7 Exponential and Logarithmic Models - College Algebra - image 3
6.7 Exponential and Logarithmic Models - College Algebra - image 3
6.7 Exponential and Logarithmic Models - College Algebra - image 4
6.7 Exponential and Logarithmic Models - College Algebra - image 4
6.7 Exponential and Logarithmic Models - College Algebra - image 5
6.7 Exponential and Logarithmic Models - College Algebra - image 5
6.7 Exponential and Logarithmic Models - College Algebra - image 6
6.7 Exponential and Logarithmic Models - College Algebra - image 6

Source Materials

Study College Algebra on Lykke

This course page is free to open: signed out, you can read its first two sections, and signing in, on any plan, opens the rest. The free Lykke Chrome extension syncs your own Canvas courses, so ChatGPT, Claude and Muse can answer from your classes and deadlines. It only reads Canvas.

Add the free extensionSign in to open every section

All sections of College Algebra · Browse more course pages

6.7 Exponential and Logarithmic Models — College Algebra | Lykke