Introduction: Why Reverse the Order of Integration?
When faced with a double integral, the first instinct is often to integrate with respect to the inner variable exactly as it appears. Changing the order can turn a complicated antiderivative into a straightforward polynomial, eliminate trigonometric substitutions, or even expose symmetries that were hidden in the original setup. Still, many integrals become much simpler after reversing the order of integration. This article explains the theory behind swapping the limits, shows step‑by‑step how to redraw the region of integration, and then works through several representative examples—culminating in a complete evaluation of a non‑trivial integral Small thing, real impact. That's the whole idea..
1. Theoretical Foundations
1.1 Fubini’s Theorem
For a bounded region (R) in the plane and a continuous function (f(x,y)) on (R), Fubini’s Theorem guarantees that
[ \iint_R f(x,y),dA = \int_{a}^{b}!\int_{g_1(x)}^{g_2(x)} f(x,y),dy,dx = \int_{c}^{d}!!In real terms, ! \int_{h_1(y)}^{h_2(y)} f(x,y),dx,dy .
The theorem tells us that the order of integration does not affect the value of the integral, provided the region and the function satisfy the required conditions (usually continuity or integrability).
1.2 Visualising the Region
The key to reversing the order is a clear picture of the region (R). Even so, in the “(dx,dy)” description, the outer limits describe how (y) varies, while the inner limits give the range of (x) for each fixed (y). To flip the order, we must rewrite those limits in terms of the opposite variable Easy to understand, harder to ignore. Surprisingly effective..
A systematic approach:
- Sketch the curves that define the original limits.
- Identify the intersection points—they become the new outer limits.
- Express the other variable as a function of the new outer variable, giving the new inner limits.
2. Step‑by‑Step Procedure
Step 1 – Write Down the Original Integral
[ I = \int_{y=a}^{y=b}\int_{x=g_1(y)}^{x=g_2(y)} f(x,y),dx,dy . ]
Step 2 – Sketch the Region
Draw the curves (x=g_1(y)) and (x=g_2(y)) together with the horizontal lines (y=a) and (y=b). Shade the region that satisfies all inequalities.
Step 3 – Determine New Outer Limits
Look at the projection of the shaded region onto the (x)-axis. The smallest and largest (x)-values become the new outer limits, say (x=c) to (x=d) Small thing, real impact..
Step 4 – Express (y) as a Function of (x)
For each fixed (x) in ([c,d]), find the lowest and highest (y) that still lie inside the region. Worth adding: these are usually obtained by solving the original boundary equations for (y). Denote them (y=h_1(x)) and (y=h_2(x)).
Step 5 – Write the Reversed Integral
[ I = \int_{x=c}^{x=d}\int_{y=h_1(x)}^{y=h_2(x)} f(x,y),dy,dx . ]
Step 6 – Evaluate
Now integrate in the new order. Often the inner integral becomes elementary, and the outer integral follows easily Took long enough..
3. Example 1 – A Simple Polynomial
[ I_1 = \int_{0}^{1}\int_{y}^{1} (x+y),dx,dy . ]
3.1 Sketch the Region
- Outer limits: (0\le y\le 1).
- Inner limits: (y\le x\le 1).
The region is the triangle bounded by (x=1), (y=0), and the line (x=y).
3.2 Reverse the Order
Projection onto the (x)-axis: (0\le x\le 1).
For a fixed (x), (y) ranges from the bottom (y=0) up to the line (y=x).
[ I_1 = \int_{0}^{1}\int_{0}^{x} (x+y),dy,dx . ]
3.3 Evaluate
Inner integral:
[ \int_{0}^{x} (x+y),dy = \Big[x y + \tfrac{1}{2}y^{2}\Big]_{0}^{x} = x^{2} + \tfrac{1}{2}x^{2} = \tfrac{3}{2}x^{2}. ]
Outer integral:
[ \int_{0}^{1} \tfrac{3}{2}x^{2},dx = \tfrac{3}{2}\Big[\tfrac{x^{3}}{3}\Big]_{0}^{1} = \tfrac{3}{2}\cdot \tfrac{1}{3}= \boxed{\tfrac{1}{2}}. ]
Notice that integrating first with respect to (x) would have required handling a quadratic in (x) inside the outer integral, a slightly messier computation. The reversed order gives a clean polynomial result.
4. Example 2 – Trigonometric Function Over a Curved Region
[ I_2 = \int_{0}^{\pi/2}\int_{0}^{\sin y} \cos x ,dx,dy . ]
4.1 Original Region
- (0\le y\le \frac{\pi}{2}).
- (0\le x\le \sin y).
The curve (x=\sin y) is the upper boundary; the region lies under the sine curve in the first quadrant.
4.2 Reverse the Order
Project onto the (x)-axis: (0\le x\le 1) (since (\sin y) reaches 1 at (y=\pi/2)).
Solve (x=\sin y) for (y): (y = \arcsin x).
For a given (x), (y) runs from the lower line (y=0) up to (y=\arcsin x).
[ I_2 = \int_{0}^{1}\int_{0}^{\arcsin x} \cos x ,dy,dx . ]
4.3 Evaluate
Inner integral is trivial because (\cos x) does not depend on (y):
[ \int_{0}^{\arcsin x} \cos x ,dy = \cos x \big[ y \big]_{0}^{\arcsin x}= \cos x , \arcsin x . ]
Now integrate with respect to (x):
[ I_2 = \int_{0}^{1} \cos x , \arcsin x ,dx . ]
Use substitution (u = \arcsin x) → (x = \sin u), (dx = \cos u , du). When (x=0), (u=0); when (x=1), (u=\pi/2).
[ I_2 = \int_{0}^{\pi/2} \cos(\sin u) , u , \cos u , du . ]
Although this looks more complicated, notice that the original integral can be evaluated directly without reversal:
[ \int_{0}^{\pi/2}!!\int_{0}^{\sin y}\cos x,dx,dy = \int_{0}^{\pi/2}\big[\sin x\big]{0}^{\sin y}dy = \int{0}^{\pi/2}\sin(\sin y),dy .
Both forms are equally challenging; the purpose of the example is to illustrate that reversal does not always simplify—the decision must be guided by the specific integrand. In many textbook problems, the reversed order leads to an elementary antiderivative, as we will see in the next example.
5. Example 3 – A Rational Function Over a Parabolic Region
[ I_3 = \int_{0}^{2}\int_{\sqrt{y}}^{2} \frac{e^{x}}{1+x^{2}} ,dx,dy . ]
5.1 Identify the Region
- Outer: (0\le y\le 2).
- Inner: (\sqrt{y}\le x\le 2).
The curve (x=\sqrt{y}) is equivalent to (y = x^{2}). The region is bounded on the left by the parabola (y=x^{2}), on the right by the vertical line (x=2), and below by (y=0) And it works..
5.2 Reverse the Order
Project onto the (x)-axis: the smallest (x) is when the parabola meets the (x)-axis, i.Plus, e. , (x=0); the largest is (x=2).
For a fixed (x) in ([0,2]), (y) runs from the bottom (y=0) up to the parabola (y = x^{2}) That alone is useful..
Thus
[ I_3 = \int_{0}^{2}\int_{0}^{x^{2}} \frac{e^{x}}{1+x^{2}} ,dy,dx . ]
5.3 Evaluate
The inner integral is immediate because the integrand does not depend on (y):
[ \int_{0}^{x^{2}} \frac{e^{x}}{1+x^{2}} ,dy = \frac{e^{x}}{1+x^{2}} , \big[ y \big]_{0}^{x^{2}} = \frac{e^{x}}{1+x^{2}} , x^{2}. ]
Hence
[ I_3 = \int_{0}^{2} \frac{x^{2} e^{x}}{1+x^{2}} ,dx . ]
Now split the fraction:
[ \frac{x^{2}}{1+x^{2}} = 1 - \frac{1}{1+x^{2}} . ]
Therefore
[ I_3 = \int_{0}^{2} e^{x},dx - \int_{0}^{2} \frac{e^{x}}{1+x^{2}},dx . ]
The first integral is elementary: (\big[e^{x}\big]_{0}^{2}=e^{2}-1).
The second integral does not have an elementary antiderivative, but it can be expressed in terms of the exponential integral ( \operatorname{Ei}) or evaluated numerically. For the purpose of this article we keep the exact representation:
[ \boxed{I_3 = e^{2}-1 - \int_{0}^{2} \frac{e^{x}}{1+x^{2}},dx }. ]
The crucial point is that after reversing the order the problem reduced to a single‑variable integral that separates into a simple part plus a well‑studied special function, whereas the original double integral would have required integrating (e^{x}/(1+x^{2})) with respect to (x) first—an impossible task without the reversal.
Honestly, this part trips people up more than it should.
6. Frequently Asked Questions
Q1: When is it safe to reverse the order?
If the region (R) is type I (vertical strips) or type II (horizontal strips) and the integrand is continuous on (R), Fubini’s theorem guarantees that the order can be swapped. For regions that are a union of such types, split the integral accordingly.
Q2: What if the region is not simply connected?
Break the region into a finite number of simple sub‑regions, reverse the order in each, and sum the results. The technique is the same; only bookkeeping becomes more involved.
Q3: Can I reverse the order for improper integrals?
Yes, provided the integral converges absolutely. Absolute convergence allows the use of Tonelli’s theorem, which extends Fubini’s result to improper cases.
Q4: Does reversing the order change the value of the integral?
No. Think about it: under the hypotheses of Fubini/Tonelli, the value remains exactly the same. Any discrepancy signals a mistake in the limits or an issue with convergence.
Q5: How do I decide which order is “better”?
Look at the integrand:
- If one variable appears only in a simple factor (e.g., (e^{x}) or (\cos y)), make that the inner variable so the inner integral collapses to a factor times the length of the interval.
- If the limits themselves are complicated functions, see whether solving them for the opposite variable yields simpler expressions.
- Consider symmetry: sometimes swapping reveals an even/odd property that leads to cancellation.
7. Practical Tips for Students
- Always sketch the region before manipulating limits. A quick drawing prevents algebraic errors.
- Label intersection points clearly; they become the new outer limits.
- When solving for the opposite variable, check domain restrictions (e.g., (\arcsin) only defined for ([-1,1])).
- After reversal, simplify the inner integral before tackling the outer one—often the inner step reduces to multiplying by the length of an interval.
- Use symbolic calculators only to verify results, not to replace the reasoning process; the goal is to understand why the reversal works.
8. Conclusion
Reversing the order of integration is a powerful technique that transforms seemingly intractable double integrals into manageable single‑variable problems. Whether the integrand is polynomial, trigonometric, or exponential, the ability to visualise the region and re‑express the limits often reveals hidden simplicity. By mastering the steps—sketching the region, finding new limits, and applying Fubini’s theorem—students and professionals alike can tackle a wide array of problems in calculus, physics, and engineering. Practice with the examples above, and soon the decision to swap the order will become an instinctive part of your problem‑solving toolbox.
Easier said than done, but still worth knowing.
