Logarithmic graphs and modelling growth and decay

Logarithmic graphs and modelling growth and decay

Spec 6.6–7 · Edexcel A-Level Maths: Revision notes

Studying this course? Keep your progress in one place.

Key facts

  • y=axny=ax^n: plot log⁡y\log y against log⁡x\log x for a straight line with gradient nn and intercept log⁡a\log a.
  • y=kbxy=kb^x: plot log⁡y\log y against xx; gradient log⁡b\log b, intercept log⁡k\log k.
  • The intercept is log⁡a\log a, so undo the log to find aa.
  • N=AektN=Ae^{kt}: k>0k>0 is growth, k<0k<0 is decay, and AA is the initial value.
  • Half-life =ln⁡2−k=\dfrac{\ln2}{-k} and doubling time =ln⁡2k=\dfrac{\ln2}{k}.
  • Evaluate a model by naming the unrealistic prediction, with numbers.

Log-log graphs

Take logarithms to get log⁡y=nlog⁡x+log⁡a\log y=n\log x+\log a, then plot log⁡y\log y against log⁡x\log x.

A log-log graph is straight for y=axny=ax^n. The gradient is nn and the vertical intercept is log⁡a\log a.

Read off both from the line, then undo the log: a=10intercepta=10^{\text{intercept}} (or eintercepte^{\text{intercept}} if you used ln⁡\ln). The intercept is log⁡a\log a, not aa.

−1−0.50.511.52−2−112345xy(0, 0.6)log y = 2.5 log x + 0.6
Log-log graph: gradient n = 2.5, intercept log a = 0.6

Worked example

A log-log graph has gradient 2.52.5 and vertical intercept 0.60.6 (using log⁡10\log_{10}). Find yy in terms of xx.

  • y=axny=ax^nlog⁡y=nlog⁡x+log⁡a\log y=n\log x+\log a

A graph of log⁡y\log y against log⁡x\log x is a straight line with intercept 11 (using log⁡10\log_{10}). What is aa?

Semi-log graphs

Take logarithms to get log⁡y=xlog⁡b+log⁡k\log y=x\log b+\log k, then plot log⁡y\log y against xx.

A semi-log graph has log⁡y\log y against xx (not log⁡x\log x). The gradient is log⁡b\log b and the intercept is log⁡k\log k, so b=10gradientb=10^{\text{gradient}} and k=10interceptk=10^{\text{intercept}}.

If the unknown is in the index plot against xx; if xx is raised to a power, plot against log⁡x\log x.

1234560.511.522.533.5xy(0, 1.2)log P = 0.30t + 1.20
Semi-log graph: gradient 0.30, intercept 1.20

Worked example

A semi-log graph of log⁡10P\log_{10}P against tt has gradient 0.300.30 and intercept 1.201.20. Find PP in terms of tt.

y = axⁿ (power law)

  • Plot log y against log x
  • Gradient n
  • Intercept log a

y = kbˣ (exponential)

  • Plot log y against x
  • Gradient log b
  • Intercept log k

Which graph gives a straight line for y=kbxy=kb^{x}?

Growth and decay

k>0k>0 gives growth without bound; k<0k<0 gives decay towards zero.

Exponential growth and exponential decay are modelled by N=AektN=Ae^{kt}, where AA is the initial value at t=0t=0. Examples: continuous compound interest, radioactive decay, drug concentration and populations. The rate of change is proportional to the current value.

Since e0=1e^0=1, N(0)=AN(0)=A. For k<0k<0, ekt→0e^{kt}\to0, so C=40e−0.25tC=40e^{-0.25t} starts at 4040 and tends to 00.

1234567891020406080100xy(0, 40)N = Ae^(kt)
N = Ae^(kt): change k to switch between growth and decay
  • N=AektN=Ae^{kt}k>0k>0 growth, k<0k<0 decay

For N=AektN=Ae^{kt} with k=−0.3k=-0.3, what happens to NN as tt becomes very large?

Finding constants

Substitute known values, isolate the exponential, then take logs.

For a half-life set N=A2N=\dfrac A2: ekt=12e^{kt}=\dfrac12, so t=ln⁡2−kt=\dfrac{\ln2}{-k}, independent of AA. For a doubling time, t=ln⁡2kt=\dfrac{\ln2}{k}.

To use a straight-line graph, take ln⁡\ln of N=AektN=Ae^{kt}: ln⁡N=ln⁡A+kt\ln N=\ln A+kt, with gradient kk and intercept ln⁡A\ln A.

12345678910−1−0.50.511.522.53tln Nintercept ln Ahalf-life: ln N falls by ln 2ln N = ln 12 − 0.3t
Taking logs of N = Ae^(kt) gives a straight line, ln N = ln A + kt (illustrative, A = 12, k = −0.3): the intercept is ln A and the gradient is k
  • Half-lifeln⁡2−k\dfrac{\ln2}{-k}
  • Doubling timeln⁡2k\dfrac{\ln2}{k}
  • ln⁡N\ln N against ttgradient kk, intercept ln⁡A\ln A

Worked example

Solve 40e−0.25t=1040e^{-0.25t}=10.

A quantity decays as N=Ae−0.1tN=Ae^{-0.1t}. Which expression gives its half-life?

Limitations of the model

Name the unrealistic prediction and say why, using numbers from the context.

Exponential growth is unbounded. P=12e0.08tP=12e^{0.08t} (in thousands) predicts about 35 80035\,800 thousand birds at t=100t=100, more than the environment can support.

A limitation is a condition where the model fails. Suggest a refinement: an upper limit, or different constants for different periods. Extrapolation beyond the fitted data is also unreliable.

51015202530354050100150200250300tP (thousands)limitP = 12e^(0.08t)with an upper limit
P = 12e^(0.08t) grows without limit; a refinement with an upper limit (illustrative, limit 120) levels off

Weak comment

  • "The model is not accurate"

Good comment

  • Names the unrealistic prediction
  • Uses numbers from the context
  • Suggests a refinement

Which is the best comment on a model that predicts unlimited growth?

Try an exam question

The concentration of a drug in the blood, CC mg/l, after tt hours is modelled by C=40e−0.25tC=40e^{-0.25t}. Find the time at which C=10C=10, to 3 significant figures, and the half-life of the drug.

[4 marks]

That's the notes covered.

Carry on to the next subtopic.

Try a question

What is the value of nn?

1 mark, marked against the mark scheme

Revision notes