14.9 Change of Variables Using Jacobians

OBJECTIVES

When you finish this section, you should be able to:

  1. Find a Jacobian in two variables (p. 963)
  2. Change the variables of a double integral using a Jacobian (p. 964)
  3. Change the variables of a triple integral using a Jacobian (p. 965)

We often find a single integral, \(\int_{a}^{b}f(x)\,{\it dx}\) using the method of substitution. To use substitution, we let \(x=g(u)\). Then \({\it dx}=g' (u)\,{\it du}, \) and \[ \int_{a}^{b}f(x)\,{\it dx}=\int_{c}^{d}f(g(u))g\prime (u)\,{\it du} \]

where \(a=g(c)\) and \(b=g(d)\). In making this substitution, the variable \(x\) is replaced by \(g(u),\) and the factor \(g\prime(u)\) is introduced into the integrand.

Sometimes it is easier to find a double integral, \(\iint\limits_{R}f(x, y) \,d\!A\) by changing the variables \(x\) and \(y\) to the polar coordinates \(r\) and \(\theta,\) by letting \(x=r\cos \theta\), \(y=r\sin \theta\), and \( d\!A=r\,dr\,d\theta\). Then \[ \begin{equation*} \iint\limits_{R}f(x,y)\,d\!A=\iint\limits_{R^{\#}}f(r\cos \theta ,r\sin \theta) r\,dr\,d\theta \end{equation*} \]

where \(R^{\#}\) is the region \(R\) expressed in polar coordinates. In changing the variables from rectangular coordinates to polar coordinates, the factor \( r\) is introduced into the integrand on the right.

When finding a triple integral \(\iiint\limits_{E}f(x, y, z)\,{dV}\), we discussed two types of change of variables: Sometimes the variables \(x, y\), and \(z\) were changed to cylindrical coordinates; other times the variables were changed to spherical coordinates.

963

Using cylindrical coordinates \((r,\theta,z),\) we get \[ \iiint\limits_{\kern-3ptE}f(x,y,z)\,{\it dV}=\iiint\limits_{\kern-3ptE^{\#}}f(r\cos \theta ,r\sin \theta ,z)~r\,dr\,d\theta \,{\it dz} \]

where \(E^{\#}\) is the solid \(E\) expressed in cylindrical coordinates. In converting from rectangular coordinates to cylindrical coordinates, the factor \(r\) is introduced into the integrand on the right.

Using spherical coordinates \(( \rho ,\theta ,\phi ) \), we have \[ \iiint\limits_{\kern-3ptE}f(x,y,z)\,{\it dV}=\iiint\limits_{\kern-3ptE^{\#}}f(\rho \sin \phi \cos \theta ,\rho \sin \phi \sin \theta ,\rho \cos \phi )\rho ^{2}\sin \phi \,d\rho \,d\theta \,d\phi\ \]

where \(E^{\#}\) is the solid \(E\) expressed in spherical coordinates. In this conversion, a factor of \(\rho ^{2}\sin \phi \) is introduced into the integrand on the right.

NOTE

In making a change in variables, the functions \(g_1\) and \(g_2\) are one-to-one.

In general, when the variables of an integral are changed, a factor, called the Jacobian, is introduced into the resulting integrand.

spanDEFINITIONspan Jacobian: Two Variables

For a change of variables from \(x,\) \(y\) to \(u,\) \(v\) given by \[\bbox[5px, border:1px solid black, #F9F7ED]{ x=g_{1}(u,v)\qquad y=g_{2}(u,v) } \]

the Jacobian of \(x\) and \(y\) with respect to \(u\) and \(v\) is \[\bbox[5px, border:1px solid black, #F9F7ED]{ \dfrac{\partial (x,y)}{\partial (u,v)} =\left\vert \begin{array}{c@{\quad}c} \dfrac{\partial x}{\partial u} & \dfrac{\partial x}{\partial v} \\[9pt] \dfrac{\partial y}{\partial u} & \dfrac{\partial y}{\partial v} \end{array} \right\vert =\dfrac{\partial x}{\partial u}\dfrac{\partial y}{\partial v}- \dfrac{\partial x}{\partial v}\dfrac{\partial y}{\partial u} } \]

NEED TO REVIEW?

Determinants are discussed in Section 10.5, p. 725.

1 Find a Jacobian in Two Variables

ORIGINS

Jacobians are named after the German mathematician Carl Gustav Jacobi (1804–1851). Jacobi was a child prodigy. By age 12, he had met the entrance requirements for university admission. He studied mathematics, Latin, and Greek and became a professor at the University of Köorionigsberg. Although he worked with functional determinants, he is most noted for his work with elliptic functions. Jacobi died of smallpox at age 46.

Finding a Jacobian

Show that in changing from rectangular coordinates \((x,y)\) to polar coordinates \((r,\theta )\), the Jacobian of \(x\) and \(y\) with respect to \(r\) and \(\theta \) is \(r\).

Solution To change the variables from rectangular to polar coordinates, we use the equations \begin{equation*} x=r\cos \theta \qquad y=r\sin \theta \end{equation*}

The partial derivatives are \[ \dfrac{\partial x}{\partial r}=\cos \theta \qquad \dfrac{\partial x}{ \partial \theta }=-r\sin \theta \qquad \dfrac{\partial y}{\partial r}=\sin \theta \qquad \dfrac{\partial y}{\partial \theta }=r\cos \theta \]

The Jacobian of \(x,y\) with respect to \(r,\theta \) is \begin{eqnarray*} \dfrac{\partial (x,y)}{\partial (r,\theta )}&=&\left\vert \begin{array}{c@{\quad}c} \dfrac{\partial x}{\partial r} & \dfrac{\partial x}{\partial \theta } \\[9pt] \dfrac{\partial y}{\partial r} & \dfrac{\partial y}{\partial \theta } \end{array} \right\vert =\left\vert \begin{array}{l@{\quad}r} \cos \theta & -r\sin \theta \\[3pt] \sin \theta & r\cos \theta \end{array} \right\vert =\cos \theta \cdot r\cos \theta -\sin \theta (-r\sin \theta)\\[4pt] &=&r\cos ^{2}\theta +r\sin ^{2}\theta =r \end{eqnarray*}

NOW WORK

Problem 9.

2 Change the Variables of a Double Integral Using a Jacobian

964

In changing variables to find a double integral, we make use of the following result, which we state without proof:

THEOREM Change in the Variables in a Double Integral

Suppose \(R\) is a closed, bounded region of the \(xy\)-plane and \(R^{\#}\) is a closed, bounded region of the \(uv\)-plane, where each point in \(R\) corresponds to one and only one point in \(R^{\#}\) when the change of variables \(x=g_{1}(u,v)\), \(y=g_{2}(u,v)\) is made. If:

  • \(f\) is continuous on \(R,\)
  • \(g_{1}\) and \(g_{2}\) have continuous partial derivatives on \(R^{\#}\), and
  • the Jacobian \(\dfrac{\partial (x,y)}{\partial (u,v)}\neq 0\) on \(R^{\#}\),

then \[\bbox[5px, border:1px solid black, #F9F7ED]{ \displaystyle\iint\limits_{\kern-3ptR}f(x,y)\,d\!A=\displaystyle\iint\limits_{\kern-3ptR^{ \#}}f(g_{1}(u,v),g_{2}(u,v))\left\vert \dfrac{\partial (x,y)}{\partial (u,v)} \right\vert \,du\,dv } \]

Notice that in the integral on the right, the absolute value of the Jacobian is introduced as a factor.

Changing variables can simplify the integration of \(\iint\limits_{\kern-3ptR}f(x,y) \,d\!A\) in two ways: either by simplifying the region \(R\) or by simplifying the integrand \(f(x,y)\).

Using a Jacobian in a Double Integral

Find \(\iint\limits_{\kern-3ptR}y\,d\!A\), where \(R\) is the region enclosed by the lines \(2x+y=0\), \(\ \ 2x+y=3\), \(x-y=0\), and \(x-y=2\).

Solution We use the change of variables \(u=2x+y\) and \(v=x-y\). Then the lines \(2x+y=0\) and \(2x+y=3\) in the \(xy\)-plane correspond to the lines \(u=0\) and \(u=3\) in the \(uv\)-plane. The lines \(x-y=0\) and \(x-y=2\) in the \(xy\)-plane correspond to the lines \(v=0\) and \(v=2\) in the \(uv\)-plane. As Figure 63 shows, the region \(R\) in the \(xy\)-plane is a parallelogram and the region \(R^{\#}\) in the \(uv\)-plane is a rectangle.

Figure 63

To find the Jacobian, we need to express \(x\) and \(y\) as functions of \(u\) and \(v\). We can achieve this by solving the system of equations \(\left\{ \begin{array}{l} u=2x+y \\ v=x-y \end{array} \right. \) for \(x\) and \(y.\) Then \[ x=\dfrac{u+v}{3} \qquad y=\dfrac{u-2v}{3} \]

The Jacobian is \begin{equation*} \dfrac{\partial (x,y)}{\partial (u,v)}=\left\vert \begin{array}{c@{\quad}c} \dfrac{\partial x}{\partial u} & \dfrac{\partial x}{\partial v}\\[3pt] \dfrac{\partial y}{\partial u} & \dfrac{\partial y}{\partial v} \end{array} \right\vert =\left\vert \begin{array}{l@{\quad}r} \dfrac{1}{3} & \dfrac{1}{3}\\[3pt] \dfrac{1}{3} & -\dfrac{2}{3} \end{array} \right\vert =-\dfrac{2}{9}-\dfrac{1}{9}=-\dfrac{1}{3} \end{equation*}

Using this change of variables, the integral \(\iint\limits_{\kern-2ptR}y\,d\!A\) becomes \[ \begin{eqnarray*} \iint\limits_{\kern-3ptR}y\,d\!A &=&\displaystyle\iint\limits_{\kern-3ptR^{\#}}\dfrac{u-2v}{3}\cdot \left\vert -\dfrac{1}{3}\right\vert \,du\,dv=\int_{0}^{2}\int_{0}^{3}\dfrac{1}{9}(u-2v)\,du\,dv=\dfrac{1}{9}\int_{0}^{2}\left[ \dfrac{u^{2}}{2}-2uv\right]_{0}^{3}\,dv\\ &=&\dfrac{1}{9}\int_{0}^{2}\left( \dfrac{9}{2}-6v\right) {dv}=\dfrac{1}{9}\left[\dfrac{9v}{2}-3v^{2}\right] _{0}^{2}=\dfrac{1}{9}(9-12)=-\dfrac{1}{3} \end{eqnarray*} \]

NOW WORK

Problem 21.

965

Using a Jacobian in a Double Integral

Find \begin{equation*} \displaystyle\iint\limits_{\kern-3ptR}(x+y)\sin (x-y)\,d\!A \end{equation*}

where \(R\) is the region enclosed by \(y=x\), \(\ y=x-2\), \(\ y=-x\), and \(\ y=-x+1 \).

Figure 64

Solution The integrand \((x+y)\sin (x-y)\) suggests changing the variables to \(u=x+y\) and \(v=x-y\). Then the lines \(x-y=0\), \(\ x-y=2\), \(\ x+y=0\), and \(\ x+y=1\) in the \(xy\)-plane define the region \(R\), and the lines \(v=0,\) \(v=2,\) \(u=0,\) and \(u=1\) in the \(uv\)-plane define the region \(R^{\#}\). See Figure 64. We solve the system of equations \(\left\{ \begin{array}{l} u=x+y \\[3pt] v=x-y \end{array} \right. \) for \(x\) and \(y\), and obtain \begin{equation*} x=\dfrac{u+v}{2}\qquad y=\dfrac{u-v}{2} \end{equation*}

Using these equations, the Jacobian is \begin{equation*} \dfrac{\partial (x,y)}{\partial (u,v)}=\left\vert \begin{array}{l@{\quad}r} \dfrac{1}{2} & \dfrac{1}{2} \\[9pt] \dfrac{1}{2} & -\dfrac{1}{2} \end{array} \right\vert =-\dfrac{1}{4}-\dfrac{1}{4}=-\dfrac{1}{2} \end{equation*}

The integral under this change of variables becomes \begin{eqnarray*} \iint\limits_{\kern-3ptR}(x+y)\sin (x-y)\,d\!A& =&\iint\limits_{\kern-3ptR^{\#}}u\sin v\cdot \left\vert -\dfrac{1}{2}\right\vert \,du\,dv=\dfrac{1}{2}\int_{0}^{1} \int_{0}^{2}u\sin v\,dv\,du \notag \\ & =&\dfrac{1}{2}\int_{0}^{1}\big[u(-\cos v)\big] _{0}^{2}\,du= \dfrac{1}{2}(1-\cos 2)\int_{0}^{1}u\,du \notag \\[4.5pt] & =&\dfrac{1}{4}(1-\cos 2) \end{eqnarray*}

NOW WORK

Problem 23.

3 Change the Variables of a Triple Integral Using a Jacobian

In space, the Jacobian is defined as follows.

spanDEFINITIONspan Jacobian: Three Variables

For a change of variables from \(x,\) \(y,\) \(z\) to \(u,\) \(v,\) \(w\) given by \[\bbox[5px, border:1px solid black, #F9F7ED]{ x=g_{1}(u,v,w)\qquad y=g_{2}(u,v,w)\qquad z=g_{3}(u,v,w) } \]

the Jacobian of \(x,\) \(y,\) and \(z\) with respect to \(u,\) \(v,\) and \(w\) is given by \[\bbox[5px, border:1px solid black, #F9F7ED]{ \dfrac{\partial (x,y,z)}{ \partial (u,v,w)}=\left\vert \begin{array}{c@{\quad}c@{\quad}c} \dfrac{\partial x}{\partial u} & \dfrac{\partial x}{\partial v} & \dfrac{ \partial x}{\partial w} \\[9pt] \dfrac{\partial y}{\partial u} & \dfrac{\partial y}{\partial v} & \dfrac{ \partial y}{\partial w} \\[9pt] \dfrac{\partial z}{\partial u} & \dfrac{\partial z}{\partial v} & \dfrac{ \partial z}{\partial w} \end{array} \right\vert } \]

NOW WORK

Problem 13.

966

In changing the variables of a triple integral, we use the following result:

THEOREM Change in the Variables in a Triple Integral

Suppose \(E\) is a closed, bounded solid of \(xyz\)-space and \(E^{\#}\) is a closed, bounded solid of \(uvw\)-space, where each point in \(E\) corresponds to one and only one point in \(E^{\#}\) when the change of variables \(x=g_{1}(u,v,w)\), \(\ y=g_{2}(u,v,w),\) and \(z=g_{3}(u,v,w)\) is made. If:

  • \(f\) is continuous on \(E\),
  • \(g_{1},\) \(g_{2},\) and \(g_{3}\) have continuous partial derivatives on \( E^{\#}\), and
  • the Jacobian \(\dfrac{\partial (x,y,z)}{\partial (u,v,w)}\neq 0\) on \( E^{\#}\), then

\[\bbox[5px, border:1px solid black, #F9F7ED]{ \iiint\limits_{\kern-3ptE}f(x,y,z)\,{\it dV}=\iiint\limits_{\kern-3ptE^{ \#}}f(g_{1}(u,v,w),g_{2}(u,v,w),g_{3}(u,v,w))\left\vert \dfrac{\partial (x,y,z)}{\partial (u,v,w)}\right\vert \,du\,dv\,dw } \]

Again notice that in the integral on the right, the absolute value of the Jacobian is introduced as a factor.

Using a Jacobian in a Triple Integral

Find \(\iiint\limits_{\kern-3ptE} y\,{\it dV},\) where \(E\) is the solid enclosed by the planes \(2x+y=0,\) \(2x+y=3,\) \(x-z=0,\) \(x-z=2\), \(z=0,\) and \(z=4.\)

Figure 65

Solution The solid \(E\) is the slanted parallelepiped shown in Figure 65(a). The change of variables \(u=2x+y,\) \(v=x-z,\) and \(w=z\) transforms \(E\) into the rectangular parallelepiped \(E^{\#}\) shown in Figure 65(b). \(E^{\#}\) is the solid enclosed by the planes \(u=0,\) \(u=3,\) \(v=0\), \( v=2,\) and \(w=0,\) \(w=4.\)

We solve the system of equations \(\left\{ \begin{array}{l} u=2x+y \\ v=x-z \\ w=z \end{array} \right. \) for \(x,\) \(y,\) and \(z\) and obtain \[ x=v+w \qquad y=u-2x=u-2 ( v+w ) =u-2v-2w \qquad z=w \]

The Jacobian is \begin{equation*} \dfrac{\partial (x,y,z)}{\partial (u,v,w)}=\left\vert \begin{array}{c@{\quad}c@{\quad}c} \dfrac{\partial x}{\partial u} & \dfrac{\partial x}{\partial v} & \dfrac{ \partial x}{\partial w} \\[9pt] \dfrac{\partial y}{\partial u} & \dfrac{\partial y}{\partial v} & \dfrac{ \partial y}{\partial w} \\[9pt] \dfrac{\partial z}{\partial u} & \dfrac{\partial z}{\partial v} & \dfrac{ \partial z}{\partial w} \end{array} \right\vert = \left|\begin{array}{c@{\quad}c@{\quad}c} 0 & \hphantom{-}1 & \hphantom{-}1\\ 1 & -2 & -2\\ 0 & \hphantom{-}0 & \hphantom{-}1 \end{array}\right| =-1 \end{equation*}

The integral \(\iiint\limits_{\kern-3ptE}\,y\,{\it dV}\) under this change of variables becomes \[ \begin{array}{rcl} \displaystyle\iiint\limits_{\kern-3ptE}~y\,{\it dV} &=&\displaystyle\iiint\limits_{\kern-3ptE^{\#}}(u-2v-2w) \left\vert \dfrac{\partial (x,y,z)}{\partial (u,v,w)}\right\vert du\, dv\, dw=\int_{0}^{4}\int_{0}^{2}\int_{0}^{3}( u-2v-2w) \,du\,dv\,dw \notag\\ &=&\displaystyle\int_{0}^{4}\int_{0}^{2}\left[ \dfrac{u^{2}}{2}-2vu-2wu\right] _{0}^{3}dv\,dw=\int_{0}^{4}\int_{0}^{2}\left( \dfrac{9}{2}-6v-6w\right) dv\, dw \notag\\ &=&\displaystyle\int_{0}^{4}\left[ \dfrac{9}{2}v-3v^{2}-6wv\right] _{0}^{2}\, dw= \int_{0}^{4}( 9-12-12w) \, dw=\int_{0}^{4}( -3-12w) \, dw \notag\\ &=&\big[ -3w-6w^{2}\big] _{0}^{4}=-12-96=-108 \end{array} \]

NOW WORK

Problem 25.