In one variable, x = g ( u ) x = g(u) x = g ( u ) gives d x = g ′ ( u ) d u dx = g'(u)\,du d x = g ′ ( u ) d u : the factor g ′ ( u ) g'(u) g ′ ( u ) says how much
a small interval stretches. In two variables, a substitution
x = x ( u , v ) x = x(u, v) x = x ( u , v ) , y = y ( u , v ) y = y(u, v) y = y ( u , v ) stretches small areas, and the stretch factor is the
absolute value of the Jacobian :
J = ∂ ( x , y ) ∂ ( u , v ) = det [ x u x v y u y v ] = x u y v − x v y u J = \frac{\partial(x, y)}{\partial(u, v)} = \det\begin{bmatrix} x_u & x_v\\ y_u & y_v \end{bmatrix} = x_u y_v - x_v y_u J = ∂ ( u , v ) ∂ ( x , y ) = det [ x u y u x v y v ] = x u y v − x v y u
The change of variables formula is then
∬ R f ( x , y ) d x d y = ∬ S f ( x ( u , v ) , y ( u , v ) ) ∣ J ∣ d u d v \iint_R f(x, y)\,dx\,dy = \iint_S f\bigl(x(u, v), y(u, v)\bigr)\,|J|\,du\,dv ∬ R f ( x , y ) d x d y = ∬ S f ( x ( u , v ) , y ( u , v ) ) ∣ J ∣ d u d v
where S S S is the region in the u v uv uv -plane that maps onto R R R .
Near a point, the map is close to linear. The small rectangle with sides d u du d u and d v dv d v
goes to a small parallelogram with sides ⟨ x u , y u ⟩ d u \langle x_u, y_u\rangle du ⟨ x u , y u ⟩ d u and
⟨ x v , y v ⟩ d v \langle x_v, y_v\rangle dv ⟨ x v , y v ⟩ d v . The area of a parallelogram spanned by two vectors is the
absolute value of their determinant . So each tiny
piece of area is multiplied by ∣ J ∣ |J| ∣ J ∣ , and the integral adds up those stretched
pieces.
For polar coordinates, x = r cos θ x = r\cos\theta x = r cos θ , y = r sin θ y = r\sin\theta y = r sin θ :
J = det [ cos θ − r sin θ sin θ r cos θ ] = r J = \det\begin{bmatrix} \cos\theta & -r\sin\theta\\ \sin\theta & r\cos\theta \end{bmatrix} = r J = det [ cos θ sin θ − r sin θ r cos θ ] = r
which is the r r r in r d r d θ r\,dr\,d\theta r d r d θ . The same calculation in three variables gives
ρ 2 sin ϕ \rho^2\sin\phi ρ 2 sin ϕ for spherical coordinates.
Example 1 — A tilted square
Evaluate ∬ R ( x + y ) e x − y d A \iint_R (x + y)\,e^{x - y}\,dA ∬ R ( x + y ) e x − y d A , where R R R is bounded by x − y = 0 x - y = 0 x − y = 0 , x − y = 2 x - y = 2 x − y = 2 ,
x + y = 1 x + y = 1 x + y = 1 and x + y = 3 x + y = 3 x + y = 3 .
Let u = x + y u = x + y u = x + y and v = x − y v = x - y v = x − y . Then R R R becomes the rectangle 1 ≤ u ≤ 3 1 \le u \le 3 1 ≤ u ≤ 3 , 0 ≤ v ≤ 2 0 \le v \le 2 0 ≤ v ≤ 2 .
Solving, x = u + v 2 x = \tfrac{u + v}{2} x = 2 u + v and y = u − v 2 y = \tfrac{u - v}{2} y = 2 u − v , so
J = det [ 1 2 1 2 1 2 − 1 2 ] = − 1 2 , ∣ J ∣ = 1 2 J = \det\begin{bmatrix} \tfrac{1}{2} & \tfrac{1}{2}\\ \tfrac{1}{2} & -\tfrac{1}{2} \end{bmatrix} = -\tfrac{1}{2}, \qquad |J| = \tfrac{1}{2} J = det [ 2 1 2 1 2 1 − 2 1 ] = − 2 1 , ∣ J ∣ = 2 1 The integral is 1 2 ∫ 1 3 u d u ∫ 0 2 e v d v = 1 2 ⋅ 4 ⋅ ( e 2 − 1 ) = 2 ( e 2 − 1 ) ≈ 12.78 \tfrac{1}{2}\int_1^3 u\,du \int_0^2 e^v\,dv = \tfrac{1}{2} \cdot 4 \cdot (e^2 - 1) = 2(e^2 - 1) \approx 12.78 2 1 ∫ 1 3 u d u ∫ 0 2 e v d v = 2 1 ⋅ 4 ⋅ ( e 2 − 1 ) = 2 ( e 2 − 1 ) ≈ 12.78 .
The region in Example 1
A square tilted 45 degrees with corners (0.5, 0.5), (1.5, 1.5), (2.5, 0.5) and (1.5, −0.5). Its sides lie on the dashed lines x − y = 0, x − y = 2, x + y = 1 and x + y = 3. In the variables u = x + y and v = x − y it becomes a rectangle.
R
1 2 3 −1 1 2 x y
x − y = 0 and x − y = 2 x + y = 1 and x + y = 3
The region in Example 1
Example 2 — The area of an ellipse
Find the area inside x 2 9 + y 2 4 = 1 \tfrac{x^2}{9} + \tfrac{y^2}{4} = 1 9 x 2 + 4 y 2 = 1 .
Let x = 3 u x = 3u x = 3 u , y = 2 v y = 2v y = 2 v . The ellipse becomes the unit disk u 2 + v 2 ≤ 1 u^2 + v^2 \le 1 u 2 + v 2 ≤ 1 , and
J = 3 ⋅ 2 = 6 J = 3 \cdot 2 = 6 J = 3 ⋅ 2 = 6 . So the area is 6 × ( area of the unit disk ) = 6 π 6 \times (\text{area of the unit disk}) = 6\pi 6 × ( area of the unit disk ) = 6 π . In general
an ellipse has area π a b \pi ab π ab .
Example 3 — Curved boundaries
Find the area of the region in the first quadrant between x y = 1 xy = 1 x y = 1 , x y = 4 xy = 4 x y = 4 , y = x y = x y = x
and y = 4 x y = 4x y = 4 x .
Let u = x y u = xy u = x y and v = y x v = \tfrac{y}{x} v = x y , so the region is 1 ≤ u ≤ 4 1 \le u \le 4 1 ≤ u ≤ 4 , 1 ≤ v ≤ 4 1 \le v \le 4 1 ≤ v ≤ 4 . It is easier to
find the Jacobian the other way:
∂ ( u , v ) ∂ ( x , y ) = det [ y x − y x 2 1 x ] = 2 y x = 2 v \frac{\partial(u, v)}{\partial(x, y)} = \det\begin{bmatrix} y & x\\ -\tfrac{y}{x^2} & \tfrac{1}{x} \end{bmatrix} = \frac{2y}{x} = 2v ∂ ( x , y ) ∂ ( u , v ) = det [ y − x 2 y x x 1 ] = x 2 y = 2 v so ∣ J ∣ = 1 2 v |J| = \tfrac{1}{2v} ∣ J ∣ = 2 v 1 . The area is ∫ 1 4 ∫ 1 4 1 2 v d v d u = 3 ⋅ ln 4 2 = 3 ln 2 ≈ 2.08 \int_1^4\int_1^4 \tfrac{1}{2v}\,dv\,du = 3 \cdot \tfrac{\ln 4}{2} = 3\ln 2 \approx 2.08 ∫ 1 4 ∫ 1 4 2 v 1 d v d u = 3 ⋅ 2 l n 4 = 3 ln 2 ≈ 2.08 .
Common mistake
Dropping the absolute value. In Example 1, J = − 1 2 J = -\tfrac{1}{2} J = − 2 1 , but area factors are
positive. Use ∣ J ∣ = 1 2 |J| = \tfrac{1}{2} ∣ J ∣ = 2 1 .
Common mistake
Using ∂ ( u , v ) / ∂ ( x , y ) \partial(u, v)/\partial(x, y) ∂ ( u , v ) / ∂ ( x , y ) directly. The formula needs ∂ ( x , y ) / ∂ ( u , v ) \partial(x, y)/\partial(u, v) ∂ ( x , y ) / ∂ ( u , v ) . If you
computed the other one, take its reciprocal, as in Example 3.
Common mistake
Changing the integrand but not the region. The limits must describe S S S in the
u v uv uv -plane. Map each boundary curve of R R R to its new equation.
Find the Jacobian of x = 2 u + v x = 2u + v x = 2 u + v , y = u − 3 v y = u - 3v y = u − 3 v .
Answer
− 7 -7 − 7
Full solution
x u y v − x v y u = 2 ( − 3 ) − 1 ⋅ 1 x_u y_v - x_v y_u = 2(-3) - 1 \cdot 1 x u y v − x v y u = 2 ( − 3 ) − 1 ⋅ 1 .
Verify that the Jacobian of polar coordinates is r r r .
Answer
r cos 2 θ + r sin 2 θ = r r\cos^2\theta + r\sin^2\theta = r r cos 2 θ + r sin 2 θ = r
Full solution
x r = cos θ x_r = \cos\theta x r = cos θ , x θ = − r sin θ x_\theta = -r\sin\theta x θ = − r sin θ , y r = sin θ y_r = \sin\theta y r = sin θ , y θ = r cos θ y_\theta = r\cos\theta y θ = r cos θ .
Find the area of the ellipse x 2 25 + y 2 16 ≤ 1 \tfrac{x^2}{25} + \tfrac{y^2}{16} \le 1 25 x 2 + 16 y 2 ≤ 1 .
Answer
20 π 20\pi 20 π
Full solution
x = 5 u x = 5u x = 5 u , y = 4 v y = 4v y = 4 v gives ∣ J ∣ = 20 |J| = 20 ∣ J ∣ = 20 times the unit disk.
Evaluate ∬ R ( x − y ) 2 d A \iint_R (x - y)^2\,dA ∬ R ( x − y ) 2 d A over the square ∣ x + y ∣ ≤ 1 \lvert x + y\rvert \le 1 ∣ x + y ∣ ≤ 1 , ∣ x − y ∣ ≤ 1 \lvert x - y\rvert \le 1 ∣ x − y ∣ ≤ 1 .
Answer
2 3 \tfrac{2}{3} 3 2
Full solution
With u = x + y u = x + y u = x + y , v = x − y v = x - y v = x − y and ∣ J ∣ = 1 2 |J| = \tfrac{1}{2} ∣ J ∣ = 2 1 : 1 2 ∫ − 1 1 d u ∫ − 1 1 v 2 d v = 1 2 ⋅ 2 ⋅ 2 3 \tfrac{1}{2}\int_{-1}^1 du\int_{-1}^1 v^2\,dv = \tfrac{1}{2} \cdot 2 \cdot \tfrac{2}{3} 2 1 ∫ − 1 1 d u ∫ − 1 1 v 2 d v = 2 1 ⋅ 2 ⋅ 3 2 .
A linear map x = a u + b v x = au + bv x = a u + b v , y = c u + d v y = cu + dv y = c u + d v sends the unit square to a parallelogram. What is its area?
Answer
∣ a d − b c ∣ \lvert ad - bc\rvert ∣ a d − b c ∣
Full solution
The Jacobian is constant, a d − b c ad - bc a d − b c , and the unit square has area 1 1 1 .
In Example 3, evaluate ∬ R x y d A \iint_R xy\,dA ∬ R x y d A instead of the area.
Answer
15 2 ln 2 ≈ 5.20 \tfrac{15}{2}\ln 2 \approx 5.20 2 15 ln 2 ≈ 5.20
Full solution
∫ 1 4 u d u ∫ 1 4 1 2 v d v = 15 2 ⋅ ln 2 \int_1^4 u\,du \int_1^4 \tfrac{1}{2v}\,dv = \tfrac{15}{2} \cdot \ln 2 ∫ 1 4 u d u ∫ 1 4 2 v 1 d v = 2 15 ⋅ ln 2 .
For u = x + 2 y u = x + 2y u = x + 2 y , v = 3 x − y v = 3x - y v = 3 x − y , find ∂ ( x , y ) / ∂ ( u , v ) \partial(x, y)/\partial(u, v) ∂ ( x , y ) / ∂ ( u , v ) .
Answer
− 1 7 -\tfrac{1}{7} − 7 1
Full solution
∂ ( u , v ) ∂ ( x , y ) = 1 ( − 1 ) − 2 ⋅ 3 = − 7 \tfrac{\partial(u, v)}{\partial(x, y)} = 1(-1) - 2 \cdot 3 = -7 ∂ ( x , y ) ∂ ( u , v ) = 1 ( − 1 ) − 2 ⋅ 3 = − 7 ; take the reciprocal.
A student evaluates Example 1 with J = − 1 2 J = -\tfrac{1}{2} J = − 2 1 and gets − 2 ( e 2 − 1 ) -2(e^2 - 1) − 2 ( e 2 − 1 ) for a positive integrand. What went wrong?
Answer
The formula uses ∣ J ∣ |J| ∣ J ∣ .
Full solution
The Jacobian’s sign only records orientation. Area scales by ∣ J ∣ = 1 2 \lvert J\rvert = \tfrac{1}{2} ∣ J ∣ = 2 1 , giving 2 ( e 2 − 1 ) 2(e^2 - 1) 2 ( e 2 − 1 ) .