Solid angle of $M^2 \subset \mathbb{R}^3$, when $0 \in M^2$ and $0 \notin M^2$

76 Views Asked by At

Let $M^2 \subset \mathbb{R}^3$ be a manifold (surface, in this case), with regular boundary $\partial M$, such that $0 \notin \partial M$. In Do Carmo's book "Differential Forms and Applications", chapter $4$, question $3b$, he defines the solid angle under which $M^2$ is seen from $0$ as

\begin{align} \Omega = \int_{M^2}\omega, \end{align}

where \begin{align} \omega = \frac{x dy \wedge dz + y dz \wedge dx + z dx \wedge dy}{(x^2 + y^2 + z^2)^{\frac{3}{2}}}. \end{align} In the previous question, I've showed that if we restric the domain of $\omega$ to $M^2$, one can obtain: \begin{align} \left. \omega \right|_{M^2} = \frac{\cos \theta}{r^2} \sigma, \end{align} where $\sigma$ is the area element of $M^2$, $r$ is the distance from $0$ to a point $p \in M^2$ and $\theta$ is the positive angle from $O_p$ to the unit positive normal $N$ to $M^2$ at p.

The real question comes now: If $0 \notin M^2$, then of course $d\omega = 0$ (one can just do a calculation on the first way we describe $\omega$). Then, the book says that we can use this and Stokes Theorem to show that, in this case, $\Omega = 0$. But, the Stokes theorem states that, in this case, $$ \Omega = \int_{M^2} \omega = \int_{\partial M} \alpha, \ \text{where} \ d \alpha = \omega. $$ But how does this concludes that $\Omega = 0$? And we know that, locally, there is a $\alpha$, but as $M^2$ is not (necessarily) contractible, we can't apply the Poincaré's Lema.

The other problem is: If $0 \notin M^2$, then $\Omega = 4 \pi$. This looks suspiciously silly, because this is the surface area of the unit sphere. But when we write $$ \int_{M^2} \frac{\cos \theta}{r^2} \sigma, $$ how am I able to calculate this? Do we need to integrate on $S^2$? Is this notation this bad or I am making a huge deal of small things? How to properly calculate all of this things? I know, those are a lot of questions, sorry about that. Thanks in advance for any help!

1

There are 1 best solutions below

1
On BEST ANSWER

The problem is Do Carmo has "swapped" his notation in the middle of the problem, and you haven't noticed this (not really your fault, I don't like his notation either). If you look at the problem again, for part (b) he says $M$ is a bounded region wit regular boundary $\partial M$. This means $M$ is a compact smooth $3$-dimensional manifold with boundary in $\Bbb{R}^3$ (think for example of a solid ellipsoid, and $\partial M$ is like the boundary elliptical surface). The problem is thus the following:

Let $M\subset \Bbb{R}^3$ be a compact, $3$-dimensional smooth submanifold with boundary $\partial M$, such that $0\notin \partial M$. Prove that \begin{align} \int_{\partial M}\omega &= \begin{cases} 0 & \text{if $0\notin M$}\\ 4\pi & \text{if $0\in M$ (and hence $0\in \text{int}(M)$)} \end{cases} \end{align}

In the case where $0\notin M$, it means $M\subset \Bbb{R}^3\setminus \{0\}$, so by Stokes theorem, \begin{align} \int_{\partial M}\omega&=\int_Md\omega =\int_M0=0. \end{align} Again, note that we were able to invoke Stokes theorem since the manifold $M$ lies in $\Bbb{R}^3\setminus\{0\}$, where the 2-form $\omega$ is defined.

For the second case, you should justify why for sufficiently small $\epsilon>0$, \begin{align} \int_{\partial M}\omega&=\int_{S_{\epsilon}}\omega = \int_{S_1}\omega=4\pi \end{align} (here $S_{\epsilon}$ is the sphere of radius $\epsilon$ centered at the origin). In other words, you have to justify why $d\omega=0$ implies that we can deform the surface of integration from $\partial M$ to $S_{\epsilon}$, and then to $S_1$ (where the integral is "obviously" equal to $4\pi$). I won't write the details here, but for the intuition in one dimension lower, see the pictures I drew in this answer (and be careful with orientations).