Double integrals over general regions — Type I and II

WHY we need "general regions"

WHY this matters: almost no real region is a rectangle. Areas, volumes, masses, centres of mass, and probabilities all live over triangles, circles, parabolic caps, etc. Type I/II are the bread-and-butter method that makes all of these computable.

Type I and Type II: the two ways to slice

Figure — Double integrals over general regions — Type I and II

DERIVATION — where the iterated integral comes from

We derive Type I from first principles (Fubini + extend-by-zero).

Step 1 — Enclose. Put $D$ inside a rectangle $R=[a,b]\times[c,d]$ where $c\le g_1(x)$ and $g_2(x)\le d$ for all $x\in[a,b]$ . Why this step? We only know how to iterate over rectangles, so we work on $R$ .

Step 2 — Extend by zero. Define $F=f$ on $D$ , $F=0$ off $D$ . Then $\iint_D f\,dA = \iint_R F\,dA$ . Why? The zero region adds no volume, so the values agree.

Step 3 — Apply Fubini on the rectangle. $\iint_R F\,dA = \int_a^b\!\!\left(\int_c^d F(x,y)\,dy\right)dx.$ Why? Fubini lets us iterate over a rectangle with constant limits.

Step 4 — Collapse the inner integral. For a fixed $x$ , $F(x,y)=0$ unless $g_1(x)\le y\le g_2(x)$ . So $\int_c^d F(x,y)\,dy = \int_{g_1(x)}^{g_2(x)} f(x,y)\,dy.$ Why? Outside $[g_1,g_2]$ the integrand is literally $0$ , contributing nothing.

Step 5 — Combine: $\iint_D f\,dA = \int_a^b\int_{g_1(x)}^{g_2(x)} f(x,y)\,dy\,dx. \qquad\blacksquare$

Type II is identical with the roles of $x$ and $y$ swapped (horizontal strips).

WORKED EXAMPLES

Worked example 1 — Triangle, both ways

Integrate $f(x,y)=x+2y$ over the triangle $D$ with vertices $(0,0),(1,0),(1,1)$ (region below the line $y=x$ , with $x\le 1$ ).

As Type I (vertical strips). For fixed $x\in[0,1]$ , $y$ runs from $0$ to the line $y=x$ . Why? The bottom is the $x$ -axis ( $y=0$ ); the top is $y=x$ . $\int_0^1\int_0^x (x+2y)\,dy\,dx.$ Inner: $\int_0^x (x+2y)\,dy = \big[xy+y^2\big]_0^x = x^2+x^2 = 2x^2.$ Outer: $\int_0^1 2x^2\,dx = \tfrac{2}{3}.$

As Type II (horizontal strips). For fixed $y\in[0,1]$ , $x$ runs from the line $x=y$ to $x=1$ . Why? On a horizontal strip the left boundary is $y=x \Rightarrow x=y$ ; the right boundary is $x=1$ . $\int_0^1\int_y^1 (x+2y)\,dx\,dy.$ Inner: $\int_y^1 (x+2y)\,dx = \big[\tfrac{x^2}{2}+2yx\big]_y^1 = (\tfrac12+2y)-(\tfrac{y^2}{2}+2y^2) = \tfrac12+2y-\tfrac{5y^2}{2}.$ Outer: $\int_0^1\!\big(\tfrac12+2y-\tfrac{5y^2}{2}\big)dy = \tfrac12+1-\tfrac{5}{6} = \tfrac{2}{3}.$ ✅ Same answer — order doesn't change the value, only the difficulty.

Worked example 3 — Switching order to make it possible

Evaluate $\displaystyle \int_0^1\int_x^1 e^{y^2}\,dy\,dx.$ Why switch: $\int e^{y^2}\,dy$ has no elementary antiderivative — the given order is a dead end. The region is $\{0\le x\le 1,\ x\le y\le 1\}$ , i.e. the triangle above $y=x$ . Re-describe as Type II: for fixed $y\in[0,1]$ , $x$ runs from $0$ to $y$ . $\int_0^1\int_0^{y} e^{y^2}\,dx\,dy = \int_0^1 y\,e^{y^2}\,dy = \tfrac12\big(e^{y^2}\big)\Big|_0^1=\tfrac{e-1}{2}.$ Why this works: now the inner integral is in $x$ with $e^{y^2}$ a constant, and the outer has the friendly factor $y$ that makes $u=y^2$ substitution trivial.

COMMON MISTAKES (steel-manned)

Recall Feynman: explain to a 12-year-old

Imagine a hilly tent and you want the space under it. Slice the floor under the tent into thin strips, like a loaf of bread. For each strip you measure how tall the tent is along it and add up — that's the inner integral. Then you add up all the strips across the whole floor — that's the outer integral. If you slice the bread the long way (vertical strips) that's Type I; if you slice it the short way (horizontal strips) that's Type II. Either way you get the same amount of space; you just pick whichever slicing is easier to measure.

What is a Type I region?

D=\{a\le x\le b,\ g_1(x)\le y\le g_2(x)\}

, integrated as

\int_a^b\int_{g_1(x)}^{g_2(x)} f\,dy\,dx

(vertical strips).

What is a Type II region?

D=\{c\le y\le d,\ h_1(y)\le x\le h_2(y)\}

, integrated as

\int_c^d\int_{h_1(y)}^{h_2(y)} f\,dx\,dy

(horizontal strips).

Why must the outer integral limits be constants?

Because the outer integral must evaluate to a number; a variable in its limits would leave an undefined expression.

How do you compute the area of a region with a double integral?

\text{Area}=\iint_D 1\,dA

What "trick" defines the integral over a non-rectangular region?

Enclose

D

in a rectangle and extend

f

0

outside

D

; area outside contributes nothing.

When is switching the order of integration essential (not just convenient)?

When the inner antiderivative doesn't exist in the given order, e.g.

\int\int e^{y^2}\,dy\,dx

— swap to integrate in

x

first.

For the triangle (0,0),(1,0),(1,1), the Type I inner limits for

y

are?

From

0

x

How do you decide which curve is the upper bound for

y

Plug in a test

x

-value and compare the two curves'

y

-values; the larger is the top (

g_2

Connections

Fubini's Theorem — the engine that justifies turning a double integral into an iterated one.
Double integrals over rectangles — the simpler base case Type I/II reduces to.
Changing the order of integration — direct application; sketch then re-read limits.
Area between two curves — recovered by setting $f\equiv 1$ .
Polar coordinates double integrals — alternative when the region is circular/annular.
Centre of mass and moments — uses these integrals with weight $f=\rho(x,y)$ .
Jacobian and change of variables — generalises slicing to skewed coordinate grids.

Concept Map

Hinglish (regional understanding)

Intuition Hinglish mein samjho

Dekho, double integral ka matlab hai surface $z=f(x,y)$ ke neeche ka volume nikalna, lekin region $D$ rectangle nahi hota — woh triangle, circle ya do curves ke beech ka area ho sakta hai. Iss curved region ko handle karne ke liye hum do tareeke use karte hain: Type I aur Type II. Type I mein hum vertical patti (strip) leke chalte hain — $x$ ko fix karke $y$ ko neeche wali curve $g_1(x)$ se upar wali curve $g_2(x)$ tak chalate hain. Type II mein horizontal patti leke chalte hain — $y$ fix karke $x$ ko left curve se right curve tak.

Sabse important rule yaad rakho: inner limit variable ho sakta hai, lekin outer limit hamesha constant (number) hota hai. Agar tumhare bahar wale integral mein $x$ ya $y$ aa gaya, samajh jao galti ho gayi, kyunki final answer to ek number hona chahiye. Aur jo curve upar hai woh hi $g_2$ banega — confirm karne ke liye ek test point daal ke dekh lo (jaise $x=\tfrac12$ pe $x^2=\tfrac14$ chhota hai $x=\tfrac12$ se, to $y=x$ upar hai).

Ek aur zabardast cheez: kabhi-kabhi order badalna sirf convenience nahi, majboori ban jaati hai. Jaise $e^{y^2}$ ka antiderivative exist hi nahi karta agar tum pehle $y$ integrate karo. Tab region ka sketch banao, Type I ko Type II mein badlo, aur achanak integral solve ho jaata hai. Isliye hamesha pehle region ki picture banao, fir limits padho — ratne ki zaroorat hi nahi, sab diagram se nikal aata hai.

Double integrals over general regions — Type I and II

WHY we need "general regions"

Type I and Type II: the two ways to slice

DERIVATION — where the iterated integral comes from

WORKED EXAMPLES

COMMON MISTAKES (steel-manned)

Connections

Concept Map

Hinglish (regional understanding)

Go deeper — visual, from zero

Test yourself — Multivariable Calculus

Connections