2
$\begingroup$

Still quite unsure about the chapter on orientation. For instance the Mobius map indeed preserves orientation, and I can show this rigorously, however I am still having difficulty understanding what it means for a holomorphic function $f$ to preserve orientation over some region in the complex plane. Which is why I asked a question regarding this $2$ days ago, but did not receive an answer.

The problem I am still thinking about is, I cannot show or provide a counter argument for why there does or does not exist some holomorphic function $f$ on a $\{z:|z|\le1\}$ such that it sends the unit circle with the counter- clockwise orientation into the unit circle with the clockwise orientation?

From the comment from this question asked earlier, one did suggest the use of the Argument principle which states that:

Argument Principle: If $f$ is meromorphic on a simply connected region $D$ and $\Gamma$ is a simply closed curve in $D$ not passing through the roots $z_j$ nor the poles $p_k$ of $f,$ then $$\frac{1}{2\pi i}\oint_{\Gamma}\frac{f'(z)}{f(z)}dz=n-m$$ Where $n$ is the number of roots of $f$(including multiplicities) in $\Gamma$ and $m$ is the number of poles of $f$ (including multiplicities) inside $\Gamma.$

I am really curious to find out why we can apply this to try and provide an explanation for my problem. I would appreciate some help because I am quite confused with orientation at the moment, and I think this problem is good in the sense it will help me understand some important notions on holomorphic functions which send or does not send certain orientation on a region to a different orientation of the same region.

I would really appreciate some rigorous argument for this problem. Much help will be appreciated.

$\endgroup$
  • 1
    $\begingroup$ Minor nitpick: You're missing a '$2\pi i$' in the formula of the argument principle. $\endgroup$ – Jonatan B. Bastos Apr 21 '18 at 4:43
4
$\begingroup$

[Following Kevin's hint on your previous question]. Suppose $f$ holomorphic on the unit disk $\overline{\mathbb D}$ maps $\partial\mathbb D$ into itself reversing orientation. Consider the integral $$ \frac1{2\pi i}\int_{|z|=1}\frac{f'(z)}{f(z)}\,dz,\ \ \ \ (\star)$$ where the counterclockwise orientation is assumed. We know its value is the number of zeros of $f$ in $\mathbb D$ by the Argument Principle (notice $f$ can't have any poles in $\mathbb D$ because it is holomorphic there). In particular this integral has to be non-negative.

On the other hand, performing the change of variables $z=e^{it}$, $$ \int_{|z|=1}\frac{f'(z)}{f(z)}\,dz=\int_0^{2\pi}\frac{f'(e^{it})}{f(e^{it})}\,ie^{it}\,dt=\int_0^{2\pi}\frac{\frac{d}{dt}\big[\,f( e^{it})\big]}{f(e^{it})}\,dt.\ \ \ \ (\ast) $$ We know that the derivative of $t\mapsto f(e^{it})$ as a curve from $[0,2\pi]$ to $\mathbb R^2$ is tangent to the unit circle $\partial\mathbb D$. Moreover, since $f$ reverses orientation the direction of this tangent vector must agree with the clockwise orientation. I.e. we have the following situation: enter image description here

Therefore $\frac{d}{dt}\big[f(e^{it})\big]\big/f(e^{it})$ has constant argument $-\pi/2$. Call $M(t)$ to its modulus. From $(\ast)$, $$ \int_{|z|=1}\frac{f'(z)}{f(z)}\,dz=\int_0^{2\pi}e^{-i\pi/2}M(t)\,dt=-i\int_0^{2\pi}M(t)\,dt. $$ From here it follows that $(\star)$ equals $\frac{-1}{2\pi}\int_0^{2\pi}M(t)\,dt$. Now, this is a non-positive number so (since we already knew $(\star)$ is non-negative) it has to be $\int_0^{2\pi}M(t)\,dt=0$ and hence $M$ is (a.e. and therefore everywhere by continuity) the $0$ function. This means that $f'=0$ on $\partial\mathbb D$, so $f'=0$ on $\overline{\mathbb D}$ (by the identity theorem for holomorphic functions for example), so $f$ must be constant. (!)

$\endgroup$
  • 1
    $\begingroup$ This was as clear as of an explanation I could have hoped for. This was very helpful thank you very much ! $\endgroup$ – Aurora Borealis Apr 21 '18 at 7:44
0
$\begingroup$

Supplementing Jonatan's answer, the reason for the preservation of orientation can also be seen by algebraic means.

Again if we consider $$\frac{1}{2\pi i}\int_{|z|=1}\frac{f'(z)}{f(z)}dz$$ With the counter clockwise orientation.

We can take a parametrization of the circle, $\gamma :[a,b] \rightarrow\mathbb{C}$, and perform the following manipulation:

$$\frac{1}{2\pi i}\int_{|z|=1}\frac{f'(z)}{f(z)}dz=\frac{1}{2\pi i}\int_{a}^{b}\frac{f'(\gamma(t))}{f(\gamma(t))}\gamma'(t)dt=\frac{1}{2\pi i}\int_{a}^{b}\frac{\frac{d}{dt}(f(\gamma(t)))}{f(\gamma(t))}dt$$

Which is, in turn $$\frac{1}{2\pi i}\int_{f\circ\gamma}\frac{1}{w}dw$$

But this is simply the winding number of the curve $f\circ \gamma$ about $0$. If we assume that $f$ maps the CCW unit circle into the CW unit circle then $f\circ \gamma$ is exactly a clockwise oriented unit circle, which means our integral must evaluate to $-1$.

But that is impossible by the argument principle, since $f$ is holomorphic in the unit circle and thus has no poles to speak of.

$\endgroup$

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.