Does Order Matter in Double Integrals? Find Out Here
Does Order Matter in Double Integrals? The Short Answer
Most students expect a straight yes or no. They don't get one. The truth is: sometimes the order of integration doesn't matter, and sometimes it does. It depends on the function and the region you're working with.
If you're evaluating a double integral over a rectangular region and the function is continuous, you can integrate in any order you like. Fubini's Theorem guarantees this. But if you're outside those nice conditions, swapping the order can change everything—or break your calculation entirely.
What the Math Actually Says
Double integrals calculate the volume under a surface over a two-dimensional region. When you write:
∬R f(x,y) dA
you're summing up tiny pieces of area multiplied by function values. The order of integration determines which variable you integrate first and which limits you use.
Fubini's Theorem: Your Golden Rule
Fubini's Theorem states that if f(x,y) is continuous on a rectangular region [a,b] × [c,d], then:
∬R f(x,y) dA = ∫ab ∫cd f(x,y) dy dx = ∫cd ∫ab f(x,y) dx dy
Both orders give you the same answer. This is the comfortable case. You can pick whichever order makes your life easier.
When Things Get Messy
Outside of continuous functions on rectangles, Fubini's Theorem doesn't apply. Here's where students get burned:
- Non-continuous functions — If f has discontinuities, the theorem breaks down
- Non-rectangular regions — The theorem requires rectangular regions specifically
- Iterated integrals with infinite limits — These require careful handling
There are actual pathological functions where the two iterated integrals exist but are different. One famous example: a function where ∫(∫f dx)dy ≠ ∫(∫f dy)dx. These aren't theoretical curiosities—they exist.
Why Changing Order Still Matters (Even When It Doesn't)
Even when the final answer is the same, the difficulty of getting there can differ dramatically. Choosing the wrong order can turn a simple problem into a nightmare of algebraic manipulation.
Example: integrating x²y over a region where y ranges from 0 to x, and x ranges from 0 to 1.
Integrating with dy first: ∫01 ∫0x x²y dy dx
Inner integral gives: ∫01 x² · (x²/2) dx = ∫01 x⁴/2 dx = 1/10
Swap the order: y goes from 0 to 1, x goes from x=y to x=1
Now you're integrating: ∫01 ∫y1 x²y dx dy
Same answer. Different path. The second setup requires more care with the limits.
How to Change the Order of Integration
When you need to swap the order, follow this process:
- Identify the region from the original limits. Draw it if needed.
- Determine the new bounds by looking at the region from the other axis.
- Rewrite the integral with the new limits and reversed integration order.
- Evaluate using the new structure.
The hardest part is usually step 2. If the region is bounded by curves, you need to describe those curves in terms of the other variable.
Quick Example
Original integral: ∫02 ∫y²4 f(x,y) dx dy
The region: y ranges from 0 to 2, and for each y, x runs from y² to 4.
Swapping: x now ranges from 0 to 4. For each x, y runs from 0 to √x.
New integral: ∫04 ∫0√x f(x,y) dy dx
That's it. Same function, same region, different limits.
Order Comparison: When to Use Which
| Situation | Recommended Order | Why |
|---|---|---|
| Region described as y = f(x) | Integrate y first | Limits are immediately available |
| Region described as x = f(y) | Integrate x first | Easier limit setup |
| One variable appears nicely in integrand | Integrate that variable first | May simplify inner integral |
| Nested fractions or roots | Try both orders | One will likely be cleaner |
| Product of separate functions | Either works | Separation makes both easy |
Common Mistakes to Avoid
- Assuming order never matters — It's true for continuous functions on rectangles, not universally
- Getting the limits wrong when swapping — This is where errors happen most often
- Forgetting to adjust the region description — The geometry must match the new integral
- Not checking if the function is continuous before assuming Fubini applies
The Bottom Line
For standard calculus problems with continuous functions on rectangular regions, the order of integration does not affect the final result. You can integrate in whatever order feels manageable.
But if you're dealing with non-rectangular regions, discontinuous functions, or iterated integrals with infinite bounds, the order can absolutely matter. Some functions have one iterated integral exist while the other doesn't.
In practice, most textbook problems are set up to be order-agnostic. Your real skill is recognizing when the order affects your workload, not just your answer. Pick the order that gives you the simplest limits and the simplest integrand after the first integration.