1
$\begingroup$

I am tryingto solve the following problem:

Let $\Omega$ be an open set of $\mathbb{R}^2$ and $\omega = \omega_1\text{d}x_1 + \omega_2\text{d}x_2 \in \text{F}_1(\Omega)$. Define $$ L = \omega_2 \frac{\partial}{\partial x_1}-\omega_1\frac{\partial}{\partial x_2}. $$ Show that
$\exists \ f \in C^{\infty}(\Omega), f > 0$ satisfying $ \text{d}(f\omega) = 0$ if and only if there exists exists $u \in C^{\infty}(\Omega)$ satifying $Lu = \frac{\partial\omega_2}{\partial x_1}-\frac{\partial\omega_1}{\partial x_2}$.

Now, I did the following:

Let $f$ have the desired properties. Then, since $f$ is a $0$-form: $$ \text{d}(f\omega) = \text{d}f \wedge \omega + (-1)^0f\wedge \text{d}\omega = \text{d}f \wedge \omega + f \text{d}\omega $$ $$ \therefore \text{d}(f\omega) = 0 \Longleftrightarrow \text{d}f\wedge\omega = -f\text{d}\omega $$ But: $$ \text{d}f = \frac{\partial f}{\partial x_1}\text{d}x_1+\frac{\partial f}{\partial x_2}\text{d}x_2 $$ $$ \therefore \text{d}f\wedge\omega = \left(\omega_2\frac{\partial f}{\partial x_1} - \omega_1\frac{\partial f}{\partial x_2}\right)\text{d}x_1\wedge\text{d}x_2 = Lf\text{d}x_1\wedge\text{d}x_2 $$ On the other side: $$ -f\text{d}\omega = -f\left(\frac{\partial \omega_2}{\partial x_1} -\frac{\partial \omega_1}{\partial x_2} \right)\text{d}x_1\wedge\text{d}x_2 $$ We then end up with $$ \text{d}(f\omega) = 0 \Longleftrightarrow Lf = -f \left(\frac{\partial \omega_2}{\partial x_1} -\frac{\partial \omega_1}{\partial x_2} \right) $$ which is almost what we wanted.

Any hints on what I may have missed or on where to go from here will be the most appreciated.

Thanks in advance.

$\endgroup$
3
$\begingroup$

With $f > 0$ and

$L = \omega_2 \dfrac{\partial}{\partial x_1} - \omega_1 \dfrac{\partial}{\partial x_2}, \tag 1$

it is easy to see that

$Lf = -f( \dfrac{\partial \omega_2 }{\partial x_1} - \dfrac{\partial \omega_1 }{\partial x_2}) \tag 2$

may be written

$-f^{-1}Lf = \dfrac{\partial \omega_2 }{\partial x_1} - \dfrac{\partial \omega_1 }{\partial x_2}; \tag 3$

also, with $f > 0$ we have

$-f^{-1}Lf = -L(\ln f), \tag 4$

since $L$ only involves first derivatives. Set

$u = -\ln f; \tag 5$

then

$Lu = L(-\ln f) = -f^{-1}Lf = \dfrac{\partial \omega_2 }{\partial x_1} - \dfrac{\partial \omega_1 }{\partial x_2} \tag 6$

as desired.

$\endgroup$
  • 1
    $\begingroup$ Note that the proof would work equally well if we had $f < 0$ everywhere; just define $u = - \ln (-f)$. The condition that $f > 0$ is perhaps best interpreted as "$f$ exists and can be chosen to be positive", not "$f$ exists and must be positive." $\endgroup$ – Michael Seifert Oct 31 '17 at 20:50
  • 1
    $\begingroup$ @MichaelSeifert: duly noted, with a tip of the hat to what was literally stated. $\endgroup$ – Robert Lewis Oct 31 '17 at 20:51
2
$\begingroup$

HINT: The fact that $f>0$ should be a hint. Try using $f=e^{-u}$ when $u$ satisfies the given equation.

$\endgroup$

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.