1
$\begingroup$

How can I calculate the integral of the 2-form $w=xdx\wedge dy+xdy\wedge dz-dx\wedge dz$ on the inferior hemisphere ($z\leq 0$) of the ellipsoid $$\frac{x^2}{9}+\frac{y^2}{9}+z^2=1$$ with the orientation determined by the normal vectors which have the third component negative on this part of the ellipsoid

I know first I should parametrize the ellipsoid: $$x = 3\cos\theta\sin\phi,\; y= 3\sin\theta\sin\phi,\; z=\cos\phi$$ Where $(\theta, \phi) \in [0,2\pi]\times [\frac{\pi}{2},\frac{3\pi}{2}]$ since ($z\leq 0$). And computing \begin{align*} \mathrm{d}x &= -3\sin\theta\sin\phi\mathrm{d}\theta + 3\cos\theta\cos\phi\mathrm{d}\phi\\ \mathrm{d}y &= 3\cos\theta\sin\phi\mathrm{d}\theta + 3\sin\theta\cos\phi\mathrm{d}\phi\\ \mathrm{d}z &=-\sin\theta\mathrm{d}\phi \end{align*}

I know I should use $$\int_M \omega = \int_U \varphi^*\omega$$

But I don't know how to compute $\varphi^*\omega$

Following @cmk answer:

I calculated everything and I have now to compute the integral $$\int_0^{2\pi}\int_{\pi/2}^{\frac{3\pi}{2}}-3\sin^2\phi(9\cos\theta \cos\phi+3\cos^2\theta \sin\phi + \sin\phi) d\theta d\phi = \int_0^{2\pi}\int_{\pi/2}^{\frac{3\pi}{2}}-9\cos^2\theta \sin^3\phi d\theta d\phi=0,$$

since the integrals $$\int_0^{2\pi}\cos(x)dx=\int_0^{2\pi}\sin(x)dx=\int_0^{2\pi}\cos(x)\sin(x)dx=0.$$

$\endgroup$
  • $\begingroup$ What is $\omega$? You have only defined $w$. $\endgroup$ – user10354138 Jun 7 at 13:12
3
$\begingroup$

If $\varphi(\theta,\phi)=(3\cos\theta\sin\phi,3\sin\theta\sin\phi,\cos\phi),$ then you compute the pullback by substitution (general form given at end):

$$\varphi^*\omega=(3\cos\theta\sin\phi)d(3\cos\theta\sin\phi)\wedge d(3\sin\theta\sin\phi)+(3\cos\theta\sin\phi)d(3\sin\theta\sin\phi)\wedge d(\cos\phi)-d(3\cos\theta\sin\phi)\wedge d(\cos\phi).$$ Note that you've computed all of the $d$'s already in your answer, so you just need to substitute them in and then compute the wedge products. I'll leave that to you to try. Then, the integration is a standard multi-variable calculus computation, which you should be able to perform without issue.

In general, the pullback of a one-form $\omega=\sum\limits_{j} a_jdx^j$ in local coordinates under $\varphi$ has the form $$\varphi^*\omega=\sum\limits_{i,j} (a_i\circ \varphi)\frac{\partial \varphi^i}{\partial x^j}dx^j.$$

$\endgroup$
  • $\begingroup$ I have edited my question with what I got. Could you please check if my result is correct? It is the first time I have done such calculus and I am unsure if what I did is ok. Thank you $\endgroup$ – codingnight Jun 7 at 16:38
  • $\begingroup$ Sure, I'll read it over. $\endgroup$ – cmk Jun 7 at 17:32
  • $\begingroup$ On a glance, it looks like in your integral, the last term in parenthesis should have a $\theta$ instead of a $\phi.$ The same is true, but in reverse, for your $dz$ expression. Just to be safe, I would encourage you to check it once more yourself, it is easy to make small mistakes on these types of things! Your process looks correct. $\endgroup$ – cmk Jun 7 at 17:54
  • $\begingroup$ @codingnight to clarify, you $dz$ should've been $-\sin\phi d\phi$ (I realize that my phasing was confusing). $\endgroup$ – cmk Jun 7 at 19:19
0
$\begingroup$

The iterated integral should be $$\pm \int_{\pi/2}^\pi \int_0^{2 \pi} f(\phi, \theta) \, d\theta d\phi.$$ You have typos in $dz$ and in the integrand. To determine the sign, check the direction of the cross product of the tangent vectors given by your parametrization.

If $\omega = F^1 \, dy \wedge dz + F^2 \, dz \wedge dx + F^3 \, dx \wedge dy$, then $d\omega = \nabla \cdot \mathbf F \, dx \wedge dy \wedge dz$. Since the integral of $x \, dx \wedge dy$ over a disk in the $x y$-plane centered at the origin is zero, your integral can also be evaluated as $$\frac 1 2 \int_{x^2/9 + y^2/9 + z^2 \leq 1} dV.$$

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.