Integration over sphere in terms of integration over disk

41 Views Asked by At

In the wave equation section of evans pde we have the following enter image description here

Can anyone explain why the jacobian term is $1+D\gamma(y) $?

1

There are 1 best solutions below

0
On BEST ANSWER

Just elaborating slightly on the case where there are two independent variables. Suppose $f:\Bbb{R}^2\to\Bbb{R}$ is a given function. Then the idea is that if you have a rectangle $A=[x_0,x_0+\Delta x]\times [y_0,y_0+\Delta y]$ in the domain, then the image $f(A)$ will be a deformed/curved rectangle which we imagine as a curved rectangular surface lying over our original rectangle $A$.

Now the question is how can we approximate the area of $f(A)$. Well we approximate our surface $f(A)$ via its tangent plane at the point $(x_0,y_0)$. Then, we have the corresponding vectors \begin{align} \xi_1=\Delta x\cdot\left(1,0,\frac{\partial f}{\partial x}(x_0,y_0)\right) \quad \text{and}\quad \xi_2=\Delta y\cdot\left(0,1,\frac{\partial f}{\partial y}(x_0,y_0)\right) \end{align} which are tangent to the surface $f(A)$ (in other words, we are taking the tangent vectors $(\Delta x,0)$ and $(0,\Delta y)$ to the rectangle $A$ and looking at the corresponding tangent vectors $\xi_1,\xi_2$ on the image $f(A)$). Now, if the rectangle $A$ is small enough, then the plane spanned by these two vectors ought to approximate $f(A)$ well. So, we can approximate the area of $f(A)$ by the area of the parallelogram spanned by the two vectors $\xi_1,\xi_2$. But if you recall, the area of a parallelogram is the absolute value of the cross product: \begin{align} \text{area } f(A) &\approx \text{area spanned by $\xi_1,\xi_2$}\\ &=|\xi_1\times \xi_2|\\ &=\sqrt{1+\left(\frac{\partial f}{\partial x}(x_0,y_0)\right)^2+ \left(\frac{\partial f}{\partial y}(x_0,y_0)\right)^2}\cdot |\Delta x\Delta y| \end{align} In higher dimensions, one would of course need an analogous formula for the $k$-dimensional volume of the parallelepiped spanned by $k$ vectors (this is ultimately related to the gram matrix I mentioned in the comments). Anyway, hopefully this gives you a better intuition for why the surface area of a graph is as such.