Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I have seen a picture for $dV$ so that $dV = r^{2} \sin(\theta)\,dr\,d\theta\,d\phi$. But how can I deduce things like $dA$ and $dV$? In a simpler coordinate (not sure about the name), $dA = r \,dr\,d\phi$, again no idea how to derive it (just given to me to memorize it). So explain what is $dA$ in polar coordinates?

share|cite|improve this question
I do not know the rigorous version of it, but since this comes from a physics background, this is mostly done by visualizing. dV is then the infinitesimal change in volume, which can be visualized as the volume of an element bound by $dr$ an arc $r\sin\theta d\theta$ and another $rd\phi$. For dA it depends which surface you're taking, $dA=dr (rsin\theta d\phi)$ or $dA=dr (r d\theta)$ or $dA=(rd\theta) (r\sin\theta d\phi)$ if you keep $\theta$, $r$ fixed or $\phi$ fixed. – Please Delete Account Mar 8 '11 at 1:35
The area element $dA = r dr d\phi$ is from polar coordinates. The boundaries are the radial $dr$ and the circumferential $rd\phi$ – Ross Millikan Mar 8 '11 at 1:38
@Approximist: there is an err. I think it must be $r \sin(\theta) d\theta$, $dr$ and $r d\phi$. N.b. the term $r^{2}$. The last part about $dA$ got me confused. Could you explain it with general surfaces? – user7628 Mar 8 '11 at 1:42
just a note, if you write $dA=r dr d\phi$ then you mean $\phi$ to be the polar angle. In the formula for dV you wrote $\phi$ would be the azimuthal angle. (@user error corrected) – Please Delete Account Mar 8 '11 at 1:43
@Approximist: technically there are some very nice general expressions for the element of surface area $dA$ for an arbitrarily parametrized surface (or even a $k$-manifold in Euclidean space). The element of $k$-dimensional volume looks like $\sqrt{Det(J^tJ)}dx_1\cdots dx_k$ where $J$ is the Jacobian matrix (matrix of partial derivatives) of the function that parametrizes the manifold. $J^t$ is the transpose of $J$. – Ryan Budney Mar 8 '11 at 19:12

Completing rcollyer's answer

$$\iint_S f(x,y) dx dy = \iint_S f(x(u,v),y(u,v)) \det[J(u,v)] du dv$$

for 2-D polar coordinates $u=r$ and $v=\theta$

Find out $\det[J(r,\theta)] =r$

Therefore $dA= r dr d\theta$

So in general your area element is $dA=\det[J(r,\theta)] du dv$

Jacobians are used extensively in Statistical mechanics, but usually most physics problems come in flavors of one in three coordinate systems(cart, spherical, cylindrical) so we do not bother with the Jacobian when geometric intuition gets us the right answer for these three cases readily, which I have written in the comment.


The Jacobian is just a product of the scale factors when you change from one (orthogonal) coordinate system to another. Please note, that all this is a part of nonstandard analysis which exploits the notion of an infinitesimal. Start working from 2-D. Your area element would be a region bounded by the lines $x,x+\Delta x,y, y+\Delta y$ and your area element is $\Delta x \Delta y$. Finally take the limit in $\Delta x ,\Delta y\rightarrow 0$ and this is understood when you write $\int \int dx dy$. Repeat with polar coordinates. To locate a curve start with the constant coordinate $r$ which is a circle of radius $r$. Then draw a curve of constant $\theta$. This is a ray which goes from the origin to infinity making an angle $\theta$ with the positive x axis. Now draw another constant curve $r+\Delta r$ and $\theta + \Delta \theta$. Look at the intersection of these two. When you take the limits, this can be approximated as a rectangle, so, area is $\Delta r (r \Delta \theta)$. The factor $r$ which converts the infinitesimal change in $\theta$ ti the corresponding displacement $rd\theta$ is called the scale factor $h_\theta$. For an orthogonal system the jacobian is the product of scale factors. For spherical coordinate system it is easy to show that: $h_r=1, h_\theta=r, h_\phi=r \sin \theta $ so $J(r,\theta,\phi)=r^2\sin\theta$ For cylindrical $h_\rho=1,h_\phi=\rho,h_z=1$ etc..

Now regarding your question about the integral, consider a transformation of a cartesian coordinate $x$ to some new coordinate $\xi$, you know that

$$\int f(x) dx = \int f(x(\xi)) \frac{dx}{d\xi}d\xi$$

The jacobian is $J(\xi)=\frac{dx}{d\xi}$

For surface integrals,

$$\iint_S f(x_1,x_2) dx_1 dx_2 = \iint_S f(x_1(\xi_1,\xi_2),x_2(\xi_1,\xi_2)) \det[J(\xi_1,\xi_2)] d\xi_1 d\xi_2$$

For volume integrals

$$\iiint f(x_1,x_2,x_3) dx_1 dx_2 dx_3 = \iiint f(x_1(\xi_1,\xi_2,\xi_3),x_2(\xi_1,\xi_2,\xi_3),x_3(\xi_1,\xi_2,\xi_3)) |J(\xi_1,\xi_2,\xi_3)| d\xi_1 d\xi_2 d\xi_3$$

EDIT2 As I mentioned in the earlier comment, when you have to find the surface element in a 3D coordinate system, you have to first know what your surface is. This would depend on what actually your finction $f(x,y)$ is. Mostly in physical problems, the surface is traced out by keeping one coordinate system fixed and varying the other two (examples- sphere's surface is fixed radii and varying azimuthal and polar angles, a plane's surface is a fixed polar angle and varying azimuthal angle and radial coordinate, etc...)

share|cite|improve this answer
@Approximist: what is the integral called? Is the base transformation a flux integral in every case? I am not sure but cannot see a reason why the integral would not differ between path -, line -, surface - and flux -integral (and possibly other unknown to me in higher dim or more advanced base things). – user7628 Mar 8 '11 at 6:57
@user: see edit – Please Delete Account Mar 8 '11 at 18:59

The way to calculate this without appeal to a geometric construction is to use the Jacobian determinant. The Jacobian matrix is constructed by taking the partial derivative of the new coordinates ($\epsilon_i$) with respect to the old coordinates ($\delta_j$)

$$J_{ij} = \frac{\partial \epsilon_i}{\partial \delta_j}.$$

share|cite|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.