Lesson 1 of 8 · 14 min

Estimating area with rectangles

Estimate the area under a curve by adding the areas of thin rectangles, using left or right endpoints, and tell whether the estimate is too big or too small.

What you will be able to do

  • Split an interval [a,b][a, b] into nn strips of width δx=b−an\delta x = \frac{b - a}{n}.
  • Calculate left-endpoint and right-endpoint estimates ∑f(xi) δx\sum f(x_i)\,\delta x of the area under a curve.
  • Decide whether an estimate is an overestimate or an underestimate from the shape of the graph.

A car travels at a steady 20 m s−120\ \mathrm{m\,s^{-1}} for 5 seconds. It covers 20×5=10020 \times 5 = 100 m, which is the area of the rectangle under its speed–time graph. If the speed keeps changing, there is no single rectangle. But over a short enough time the speed hardly changes, so you can chop the time into short pieces, treat each as a rectangle, and add them up.

That idea, adding up thin rectangles, is how you find the area under any curve. It is the start of everything in this topic.

Four left-endpoint rectangles under y=x2+1y = x^2 + 1 on [0,2][0, 2]. Each height is ff at the left edge of the strip.

Definition

A sum of rectangles

Split [a,b][a, b] into nn strips of equal width

δx=b−an\delta x = \frac{b - a}{n}

Choose a point xix_i in each strip and use f(xi)f(x_i) as the height of that strip's rectangle. The total area of the rectangles is

∑if(xi) δx=f(x1) δx+f(x2) δx+⋯+f(xn) δx\sum_{i} f(x_i)\,\delta x = f(x_1)\,\delta x + f(x_2)\,\delta x + \cdots + f(x_n)\,\delta x

With left endpoints, each xix_i is the left edge of its strip; with right endpoints, the right edge. The symbol ∑\sum (capital sigma) means "add up".

Interactive

Rectangles under a curve

Switch between left and right endpoints, then increase nn. For this increasing curve, left rectangles sit under the curve and right rectangles stick out above it. Both estimates close in on 143≈4.667\frac{14}{3} \approx 4.667 as the strips get thinner.

Worked example

Left and right estimates of the area under y=x2+1y = x^2 + 1

  • Simple familiar
  • Technology free

Estimate the area under y=x2+1y = x^2 + 1 between x=0x = 0 and x=2x = 2 using 4 rectangles of equal width with

(a) left endpoints, and (b) right endpoints.

State whether each is an overestimate or an underestimate.

5 steps

Try it yourself first, then check the working one step at a time.

Method

  1. Find δx=b−an\delta x = \frac{b - a}{n} and list the strip edges.

  2. Tabulate ff at the edges.

  3. Left: add f(x0)f(x_0) to f(xn−1)f(x_{n-1}). Right: add f(x1)f(x_1) to f(xn)f(x_n). Multiply by δx\delta x.

  4. Use whether ff is increasing or decreasing to decide over or under.

Watch out for

  • Forgetting to multiply the sum of the heights by δx\delta x.

  • Using five heights for four rectangles.

The interval [2,5][2, 5] is split into 6 strips of equal width. What is the width δx\delta x of each strip?

Estimate the area under y=x2y = x^2 from x=1x = 1 to x=3x = 3 using 4 rectangles of equal width with right endpoints.

Key idea

Too big or too small?

If ff is increasing on the interval, left-endpoint rectangles lie under the curve (an underestimate) and right-endpoint rectangles poke above it (an overestimate). If ff is decreasing, it is the other way round. If ff goes up and down, look at the graph strip by strip.

Either way, thinner strips give a better estimate.

The function f(x)=e−xf(x) = e^{-x} is decreasing for all xx. Which statement about estimates of the area under y=e−xy = e^{-x} from x=0x = 0 to x=2x = 2 is true?

Choose one answer

Common mistake

Sum of heights is not an area

Add the heights, then multiply by the width. And count your endpoints: nn rectangles use nn heights. Left endpoints use x0x_0 to xn−1x_{n-1} and leave out bb; right endpoints use x1x_1 to xnx_n and leave out aa.

Exam tip

Set it out in a table

List xix_i and f(xi)f(x_i) in a small table before you add. Markers can then give you method marks for the right endpoints and heights even if an arithmetic slip creeps in. On Paper 2, a calculator list or the sum command saves time for larger nn.

Summary

  • Split [a,b][a, b] into nn strips of width δx=b−an\delta x = \frac{b - a}{n}.

  • A rectangle sum ∑f(xi) δx\sum f(x_i)\,\delta x estimates the area under the curve. Left and right endpoints give two different estimates.

  • For an increasing function, left endpoints underestimate and right endpoints overestimate; for a decreasing function it is the reverse.

  • Thinner strips (larger nn) give better estimates.

  • Multiply the sum of the heights by the strip width.

Finished the lesson?

Try the 3 quick checks above and the lesson completes itself, or mark it complete now.

Lesson 1 of 80% read

Save your progress with a free account