A student immediately sets up ∫∫ f(x,y) dx dy to compute ∬_R f dA, reasoning that a double integral is just two nested single integrals by definition. What is wrong with this reasoning?
ANothing — the double integral is defined as two nested single integrals in the order dx dy
BThe student should integrate dy dx instead, since x comes first in the Cartesian plane
CComputing as iterated integrals is justified by Fubini's theorem, not the definition — the definition is a limit of Riemann sums
DThe student needs to include a Jacobian factor to account for the 2D area element
The double integral ∬_R f dA is defined as the limit of 2D Riemann sums — the same limiting process as in 1D, extended to rectangles in the plane. Fubini's theorem then proves (under appropriate conditions) that this limit equals the iterated integral ∫∫ f dy dx or ∫∫ f dx dy. The iterated-integral computation method and the definition are different things. This distinction matters because Fubini's theorem has conditions, and it explains why you can sometimes swap integration order (justified by Fubini) but can't always do so blindly.
Question 2 Multiple Choice
If f(x,y) = 1 everywhere on region R, what does ∬_R f(x,y) dA equal?
A1, since f is constantly 1 and integration of a constant just returns that constant
B0, since a flat surface at height 1 has no interesting volume
CThe area of region R
DThe perimeter of region R
When f = 1, the Riemann sum becomes ∑ 1·ΔA = ∑ ΔA, which approximates the total area of region R. In the limit, ∬_R 1 dA = area(R). Geometrically, the 'volume' under the flat surface z = 1 above R is just the base area R times height 1. This is the simplest and most important special case: the double integral degenerates to an area computation when the integrand is 1.
Question 3 True / False
The double integral ∬_R f(x,y) dA can be negative when f takes negative values on part of R.
TTrue
FFalse
Answer: True
Just like the 1D definite integral gives signed area (subtracting where f < 0), the double integral gives signed volume. Where f < 0, the 'boxes' in the Riemann sum have negative height, contributing negatively to the sum. If f is negative everywhere on R, ∬_R f dA is negative. This makes double integrals useful for computing net quantities (like net electric charge or net probability density) rather than just positive volumes.
Question 4 True / False
The double integral ∬_R f(x,y) dA is defined as two successive single integrals; this is its definition, not a theorem.
TTrue
FFalse
Answer: False
This is the key conceptual error to avoid. The double integral is defined as the limit of 2D Riemann sums: partition R into small rectangles, evaluate f at sample points, multiply by area ΔA, sum, and take the limit as rectangle size shrinks. Fubini's theorem is what guarantees (under appropriate conditions on f and R) that this limit equals the iterated integral ∫_a^b [∫_c^d f(x,y) dy] dx. Definition and computation method are different things, and conflating them causes errors when Fubini's conditions aren't met.
Question 5 Short Answer
Why is it important to distinguish the definition of the double integral from the method of computing it as iterated integrals?
Think about your answer, then reveal below.
Model answer: The definition (limit of Riemann sums) tells you what the double integral means and when it exists. The iterated-integral computation method is justified by Fubini's theorem, which has conditions (f must be integrable, and for non-rectangular regions, the limits of integration must correctly describe the region). Keeping them separate matters in practice because: (1) when changing the order of integration, you must re-derive the limits from the geometry of R — the region doesn't change, but how you slice it does; (2) some functions violate Fubini's conditions and the two iterated integrals give different values even though the double integral is well-defined.
The practical payoff of this conceptual clarity is that you can swap integration order confidently (because Fubini justifies it) while knowing that the limits of integration must be rederived from scratch — they aren't symmetric. You also recognize that ∬ 1 dA = area(R) and ∬ f dA can be negative, which are properties of the definition, not artifacts of the computation method.