Riemann sums, sigma notation, signed area, accumulation functions, and the definite integral as total change — the geometry and meaning of integration
Chapter 9 introduced antiderivatives as the reverse process of differentiation and showed how to find a family of functions whose derivative equals a given function. That work is essential, but it is not yet the full story of integration. Antiderivatives answer the question: what function has this derivative? Definite integrals answer a different question: how much of something has accumulated over an interval?
This chapter introduces the definite integral as a precise measure of accumulated quantity. Students will learn to approximate area under a curve using rectangles, to organize those approximations using Riemann sums and sigma notation, and to represent accumulated quantities using definite integral notation. Along the way, the chapter develops the ideas of signed area, net area, and total area, and shows how integrals can measure far more than geometric area. By the end of this chapter, students will be ready for the Fundamental Theorem of Calculus, which will connect the definite integral to antiderivatives in a deep and remarkable way.
A car is traveling along a straight road. A speedometer gives its velocity at every moment in time. If you know the car's velocity at every instant from time \(t = 0\) to time \(t = 5\) seconds, can you determine how far the car has traveled? Velocity describes how fast distance is changing. Distance is the accumulated result of velocity acting over time. If velocity were constant, you could multiply velocity by time. But velocity changes continuously. How do you accumulate a quantity that changes at every moment? This is the central question of integral calculus, and the answer is the definite integral.
The question of finding the area of a region bounded by a curve and the horizontal axis seems like a geometry problem. And in a sense, it is. But it is also something more than that: it is the starting point for understanding accumulation.
Consider the graph of a function \(f(x)\) that is positive on the interval \([a, b]\). The region between the graph of \(f\) and the \(x\)-axis, from \(x = a\) to \(x = b\), has a well-defined area. For simple shapes such as rectangles, triangles, and trapezoids, formulas from geometry give the exact area. But for curved boundaries, the situation is more complicated.
The basic idea: If no formula exists for the area exactly, estimate it using shapes whose areas you can compute. Rectangles are the natural choice, because every rectangle has area equal to base times height.
If you divide the interval \([a, b]\) into several smaller subintervals and erect a rectangle on each one, then the combined area of all the rectangles approximates the area under the curve. The more rectangles you use, the thinner each one becomes, and the better the approximation.
This idea — estimating curved area with many thin rectangles and then making the rectangles thinner and thinner — is the geometric foundation of the definite integral.
Draw the graph of a smooth, positive, increasing function \(f(x)\) over the interval \([a, b]\). Divide the interval into five equal subintervals. Draw five rectangles whose bases rest on the \(x\)-axis and whose tops touch the curve at some point within each subinterval. Label the width of each rectangle as \(\Delta x\) and the height as \(f(x_i)\) for some sample point \(x_i\). The total shaded area of all five rectangles approximates the area under the curve. The key insight: this is an approximation, not an exact answer.
The exact area requires an infinite limiting process: make the rectangles infinitely thin. This limiting process leads directly to the definite integral. Approximation with rectangles is not a workaround — it is how the definite integral is actually defined.
The choice of where the rectangle height is measured within each subinterval is a design choice. Three natural choices produce three standard approximations. In each case, divide \([a, b]\) into \(n\) equal subintervals of width:
The endpoints of the subintervals are \(x_0 = a,\; x_1 = a + \Delta x,\; \ldots,\; x_n = b\).
Heights are taken from the left endpoint of each subinterval. For an increasing function, this underestimates; for a decreasing function, it overestimates.
Heights are taken from the right endpoint of each subinterval. For an increasing function, this overestimates; for a decreasing function, it underestimates.
Heights are taken from the midpoint \(\bar{x}_i = \frac{x_{i-1}+x_i}{2}\) of each subinterval. The midpoint sum is often a better approximation than either left or right sums for the same value of \(n\).
L₄ = f(0)(0.5) + f(0.5)(0.5) + f(1)(0.5) + f(1.5)(0.5)
= 0 + 0.125 + 0.5 + 1.125 = 1.75
R₄ = f(0.5)(0.5) + f(1)(0.5) + f(1.5)(0.5) + f(2)(0.5)
= 0.125 + 0.5 + 1.125 + 2 = 3.75
M₄ = f(0.25)(0.5) + f(0.75)(0.5) + f(1.25)(0.5) + f(1.75)(0.5)
= 0.03125 + 0.28125 + 0.78125 + 1.53125 = 2.625
For an increasing function on \([a, b]\): \(L_n < \text{true area} < R_n\).
For a decreasing function on \([a, b]\): \(R_n < \text{true area} < L_n\).
The midpoint sum is typically more accurate than either, but confirming requires additional analysis of the function's concavity.
The left, right, and midpoint sums are all special cases of a more general idea: the Riemann sum. Named after the mathematician Bernhard Riemann, a Riemann sum allows the sample point within each subinterval to be chosen freely. This flexibility is mathematically important because it shows that the limit of these approximations does not depend on which sample point is chosen — as long as the rectangle widths shrink to zero.
Let \(f\) be defined on \([a, b]\). Divide \([a, b]\) into \(n\) subintervals. Choose any sample point \(x_i^*\) in the \(i\)-th subinterval. With equal subinterval widths \(\Delta x = \frac{b-a}{n}\):
\(\displaystyle\sum_{i=1}^{n} f(x_i^*)\,\Delta x\)
As \(n \to \infty\) (and \(\Delta x \to 0\)), this sum approaches the exact area when \(f\) is integrable on \([a, b]\).
Left, right, and midpoint sums are all Riemann sums. All three converge to the same limit when \(f\) is continuous. The Riemann sum is not an exact answer — it is a controlled approximation whose accuracy improves as \(n\) grows. The definite integral is defined as the limit of these Riemann sums as \(n \to \infty\).
Riemann sums involve adding many similar terms. Sigma notation provides a compact, precise way to write these sums. The Greek letter \(\Sigma\) (sigma) indicates summation.
\(\displaystyle\sum_{i=1}^{n} a_i\) means: add all terms of the form \(a_i\) as \(i\) runs from 1 to \(n\).
i — the index of summation · i = 1 — starting value · n — ending value · a_i — the general term
The right-hand Riemann sum for \(f\) on \([a, b]\) with \(n\) equal subintervals in sigma notation:
R_n = ∑ f(2i/n) · (2/n) = ∑ (2i/n)² · (2/n) = (8/n³) ∑ i²
R_n = (8/n³) · n(n+1)(2n+1)/6 = 4(n+1)(2n+1)/(3n²)
lim(n→∞) 4(n+1)(2n+1)/(3n²) = 4·2/3 = 8/3
The limit of a Riemann sum as the number of subintervals grows without bound is the definite integral. When \(f\) is continuous on \([a, b]\), this limit exists and equals:
∫ — an elongated S (for sum); signals accumulation
a — the lower limit of integration; where accumulation begins
b — the upper limit of integration; where accumulation ends
f(x) — the integrand; the function being accumulated
dx — the variable of integration; corresponds to \(\Delta x\) in the Riemann sum; also carries units
Together, \(\int_a^b f(x)\,dx\) says: accumulate the values of \(f(x)\) over infinitely many infinitesimally thin slices from \(x = a\) to \(x = b\).
The definite integral \(\int_a^b f(x)\,dx\) is a number, not a function. The limits \(a\) and \(b\) fix the interval, so the result is a specific accumulated value. The variable \(x\) is a dummy variable — \(\int_a^b f(x)\,dx\), \(\int_a^b f(t)\,dt\), and \(\int_a^b f(u)\,du\) all represent the same number.
If \(f(x)\) has units of \(U_1\) and \(x\) has units of \(U_2\), then \(\int_a^b f(x)\,dx\) has units of \(U_1 \cdot U_2\). For example, if velocity is in meters per second and \(t\) is in seconds, then \(\int_a^b v(t)\,dt\) has units of meters — it measures displacement, not speed.
Definite integrals are defined for functions that may dip below the horizontal axis. When \(f(x) < 0\) on some part of \([a, b]\), the product \(f(x_i^*)\,\Delta x\) is negative, because \(f(x_i^*)\) is negative and \(\Delta x\) is positive. These negative contributions reduce the value of the Riemann sum.
This leads to the concept of signed area: area above the \(x\)-axis is counted as positive, and area below the \(x\)-axis is counted as negative.
Draw a smooth function that is positive from \(x = a\) to \(x = c\) and negative from \(x = c\) to \(x = b\). Shade the region above the axis in one color, labeled positive area \(A_1\). Shade the region below the axis in a second color, labeled negative area \(A_2\). The definite integral \(\int_a^b f(x)\,dx = A_1 - A_2\). The total unsigned area is \(A_1 + A_2\).
Signed area can be zero even when the function is not zero everywhere. If equal amounts of area fall above and below the axis, the definite integral equals zero. A zero integral does not mean the function is identically zero on \([a, b]\).
Two related but distinct quantities arise from signed area.
Net area (signed area): The value of the definite integral \(\int_a^b f(x)\,dx\), computed with sign. Net area reflects accumulation that allows positive and negative contributions to cancel.
Total area (unsigned area): The actual geometric area enclosed between the graph and the \(x\)-axis, regardless of sign.
To compute total area, split the interval at every point where \(f(x) = 0\), integrate \(f\) over each subinterval, and add the absolute values. A classic example: \(\int_0^{2\pi} \sin x\,dx = 0\) (signed areas cancel), but \(\int_0^{2\pi} |\sin x|\,dx = 4\) (total geometric area).
| Interpretation | Formula | When to Use |
|---|---|---|
| Net area | \(\int_a^b f(x)\,dx\) | Net displacement, net population change, net revenue |
| Total area | \(\int_a^b |f(x)|\,dx\) | Total distance, total blood flow, total energy expended |
Definite integrals can be evaluated geometrically when the graph of \(f\) produces recognizable shapes. No antiderivative computation is needed.
| Shape | Area Formula |
|---|---|
| Rectangle (width \(w\), height \(h\)) | \(wh\) |
| Triangle (base \(b\), height \(h\)) | \(\frac{1}{2}bh\) |
| Semicircle (radius \(r\)) | \(\frac{1}{2}\pi r^2\) |
| Trapezoid (parallel sides \(a\), \(b\), height \(h\)) | \(\frac{1}{2}(a+b)h\) |
Region 1 (triangle, above): A₁ = ½(2)(4) = 4
Region 2 (rectangle, above): A₂ = (3)(4) = 12
Region 3a (triangle, x=5 to x=7, above): A₃ₐ = ½(2)(4) = 4
Region 3b (triangle, x=7 to x=8, below): A₃ᵦ = ½(1)(2) = 1 (negative)
∫₀⁸ g(x) dx = 4 + 12 + 4 − 1 = 19
When a function is known only through a table of values, definite integrals can be estimated using Riemann sums directly from the table. Identify the subinterval widths from the \(x\)-values, choose the appropriate sum (left, right, or midpoint), multiply each function value by its corresponding \(\Delta x\), and add.
| \(t\) (min) | 0 | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|---|
| \(r(t)\) (gal/min) | 12 | 15 | 14 | 18 | 20 | 17 |
L₅ = r(0)(1) + r(1)(1) + r(2)(1) + r(3)(1) + r(4)(1)
= 12 + 15 + 14 + 18 + 20 = 79 gallons
R₅ = 15 + 14 + 18 + 20 + 17 = 84 gallons
The definite integral \(\int_a^b f(x)\,dx\) is a fixed number for fixed \(a\), \(b\), and \(f\). But what if the upper limit of integration varies? Consider the function:
where \(a\) is a fixed starting point and \(x\) is a variable upper limit. This is called an accumulation function (or area function). For each value of \(x\), \(A(x)\) gives the accumulated signed area under \(f\) from \(a\) to \(x\).
Note: the variable inside the integral is called \(t\) (a dummy variable) to avoid confusion with the upper limit \(x\).
\(A(a) = 0\) — accumulation starting and ending at the same point is zero.
\(A(x)\) is increasing when \(f(x) > 0\) and decreasing when \(f(x) < 0\).
Chapter 11 will prove: \(A'(x) = f(x)\). The rate of change of accumulation equals the current function value. This is the Fundamental Theorem of Calculus.
Draw a positive function \(f(t)\) above the \(t\)-axis. Show the shaded region from \(t = a\) to \(t = x\) for several increasing values of \(x\). As \(x\) increases, the shaded region grows, and so does \(A(x)\). Draw the graph of \(A(x)\) below, starting at \(A(a) = 0\) and increasing (with varying rate) as \(x\) increases. Where \(f(t)\) is large, \(A(x)\) grows steeply. Where \(f(t)\) is small, \(A(x)\) grows slowly.
A(x) = 2xOne of the most important interpretations of the definite integral is this: if \(f'(x)\) gives the rate of change of some quantity \(F(x)\), then \(\int_a^b f'(x)\,dx\) gives the total change in \(F\) from \(x = a\) to \(x = b\). This is the Net Change Theorem:
In words: the integral of a rate of change gives the net (total) change in the original quantity. Chapter 11 will develop this more carefully through the Fundamental Theorem.
| Rate \(f'(x)\) | Accumulated quantity \(\int_a^b f'(x)\,dx\) |
|---|---|
| Velocity (m/s) | Displacement (m) |
| Acceleration (m/s²) | Change in velocity (m/s) |
| Marginal cost (\$/item) | Change in total cost (\$) |
| Population growth rate (people/year) | Total population change (people) |
| Flow rate (L/min) | Volume accumulated (L) |
| Power (watts = J/s) | Energy (joules) |
The integral of a rate of change gives net change, not total magnitude. If a population grows and then shrinks, the integral gives the net change (possibly small), not the total growth plus the total decline separately. Distinguish carefully between what is being asked.
Motion problems provide the clearest illustrations of how definite integrals measure accumulated change. Let \(s(t)\) be position, \(v(t) = s'(t)\) velocity, and \(a(t) = v'(t)\) acceleration.
Displacement (signed): \(\displaystyle\int_a^b v(t)\,dt = s(b) - s(a)\)
Total distance traveled (unsigned): \(\displaystyle\int_a^b |v(t)|\,dt\)
These are equal only when \(v(t)\) does not change sign on \([a, b]\). When the object reverses direction, total distance exceeds the magnitude of displacement.
Displacement = ∫₀⁴ (t² − 4) dt = [t³/3 − 4t]₀⁴
= (64/3 − 16) − 0 = 64/3 − 48/3 = 16/3 ≈ 5.33 m
Distance (back): ∫₀² (4 − t²) dt = [4t − t³/3]₀² = 8 − 8/3 = 16/3 ≈ 5.33 m
Distance (forward): ∫₂⁴ (t² − 4) dt = [t³/3 − 4t]₂⁴
= (64/3 − 16) − (8/3 − 8) = 56/3 − 8 = 32/3 ≈ 10.67 m
Total distance = 16/3 + 32/3 = 48/3 = 16 m
V = ∫₀² (3 + sin(πt)) dt = [3t − cos(πt)/π]₀²
= (6 − cos(2π)/π) − (0 − cos(0)/π)
= (6 − 1/π) − (−1/π) = 6 gallons
A definite integral evaluates to a specific number. The constant of integration disappears when limits of integration are applied. Write \(\left[\frac{x^3}{3}\right]_0^3 = 9\), not \(\frac{x^3}{3} + C\).
The definite integral gives signed area automatically. If you want total unsigned area, split the integral at the zeros of \(f\) and take absolute values of each piece separately. Do not blindly make everything positive.
\(\int_a^b v(t)\,dt\) gives displacement — signed net change in position. Total distance requires \(\int_a^b |v(t)|\,dt\). These are equal only when \(v(t)\) does not change sign.
Reversing the limits changes the sign: \(\int_a^b f\,dx = -\int_b^a f\,dx\). Check which direction accumulation runs before computing.
In a left-hand sum, the height comes from the left endpoint of each subinterval. In a right-hand sum, from the right endpoint. Mixing these produces incorrect approximations.
A zero integral means positive and negative contributions cancel, not that \(f(x) = 0\). For example, \(\int_0^{2\pi} \sin x\,dx = 0\) even though \(\sin x\) is certainly not zero everywhere.
The constant of integration cancels: \([F(x)+C]_a^b = (F(b)+C)-(F(a)+C) = F(b)-F(a)\). There is no need to include it when evaluating definite integrals.
If \(r(t)\) is in liters per minute but \(t\) is in seconds, units must be reconciled before integrating. Failure to convert produces a numerically meaningless result.
This chapter has developed the idea of the definite integral through approximation, geometry, signed area, and accumulation. But evaluating definite integrals by computing Riemann sums and taking limits is tedious. The approximations from left and right sums improve as \(n\) increases, but computing exact values this way requires sophisticated algebraic work with summation formulas.
Chapter 11 introduces the Fundamental Theorem of Calculus, which provides a vastly more efficient method: evaluate a definite integral by finding an antiderivative. This result is not obvious. It says something profound — the process of differentiation (measuring instantaneous change) and the process of integration (accumulating quantity over an interval) are inverse operations. They are connected at the deepest level.
The concepts established in this chapter — Riemann sums, the definition of the definite integral, signed area, accumulation functions — form the foundation for understanding why the Fundamental Theorem works and what it means, not just how to apply it.
The accumulation functions introduced in Section 10.10 are the key. Understanding that \(A(x) = \int_a^x f(t)\,dt\) is a function with derivative \(f(x)\) is the heart of the Fundamental Theorem. Everything in Chapters 9 and 10 — antiderivatives, indefinite integrals, Riemann sums, signed area, and accumulation — has been building toward this moment. In Chapter 11, computing \(\int_a^b f(x)\,dx\) will require only finding \(F(b) - F(a)\), where \(F\) is any antiderivative of \(f\).
Evaluate each definite integral using geometric area. No antiderivative needed.
| \(t\) | 0 | 2 | 4 | 6 | 8 | 10 |
|---|---|---|---|---|---|---|
| \(r(t)\) | 5 | 8 | 6 | 9 | 7 | 4 |
Each problem shows a student's error. Identify it and write the correct solution.
Identify and correct the error. What does a zero integral actually mean?
Explain what information is actually needed to determine whether \(L_n\) overestimates or underestimates.
Is the student correct to be suspicious? Explain why a negative definite integral can be correct.
Identify the conceptual error and explain the correct interpretation.
What is wrong? Write the correct evaluation. What happens to \(C\)?
The definite integral has now been defined, interpreted, and approximated. But a fundamental question remains open: is there a method more efficient than Riemann sums for computing exact values? The answer is yes, and it is one of the most remarkable theorems in all of mathematics. Chapter 11 shows that to find \(\int_a^b f(x)\,dx\) exactly, you need only find an antiderivative \(F\) of \(f\) and compute \(F(b) - F(a)\). No Riemann sums. No limits. No sigma notation. The accumulation functions from Section 10.10 are the key: understanding that \(A'(x) = f(x)\) is the heart of the Fundamental Theorem. Derivatives and integrals are inverse processes. Everything in Chapters 9 and 10 has been building toward this moment.