MathOverflow is a question and answer site for professional mathematicians. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Suppose that $f:U\subset\mathbb{C}\to\mathbb{C}$, where $U$ is a region in the complex plane, is a holomorphic function.

Of course, if $c\in\mathbb{R}$ is a regular value for $\text{Re}(f(z))$ then $\text{Re}(f(z))^{-1}(c)$ is an union of differentiable curves in the plane.


If $c$ is not a regular value and $\text{Re}(f(z))^{-1}(c)$ have at least one cluster point is this set also a piece-wise differentiable curve ?

share|cite|improve this question
Yes, but they may intersect each other. Let us say $z$ is a critical point if $f'(z)=0$, and let us treat the set $\mathrm{Re}f(z)=c$ as the preimage of the curve $\mathrm{Re}w=c$. Then at non-critical points $f$ is locally invertible, so the curve $\mathrm{Re}w=c$ has differentiable curve as its preimage near non-critical points. Now near its critical points $f$ is $n$-to-$1$, and so the curve $\mathrm{Re}w=c$ will have as its preimage $n$ copies of differentiable curves, which all intersect (forming an even angle between them) at the critical point. – timur Nov 26 '10 at 17:17
up vote 5 down vote accepted

Yes. The set of points in $\mathbb C$ having real part equal to $c$ form a line, i.e. a smooth simple curve $l$. The counterimage $f^{-1}(l)$ of any smooth simple curve $l$ via a holomorphic function $f$ is always piecewise smooth.

To prove the last sentence, take any point $z_0\in U$ with $f(z_0) \in l$. There are two local diffeomorphisms at $z_0$ and $f(z_0)$ that move $z_0$ to $0$ and transform $f$ locally into $g(z) = z^n$. Since $f$ is not constant, we have $n>0$. Local diffeomorphisms send smooth curves to smooth curves. The counterimage along $g$ of a smooth curve passing through $0$ is the union of $n$ smooth curves exiting from 0. Therefore $f^{-1}(l)$ is a piecewise smooth curve.

share|cite|improve this answer
Hi Bruno, thank you for write this. Nice answer. – Leandro Nov 26 '10 at 23:01

You are essentially asking for a description of the zero-set of a harmonic function in two variables. That is a very delicate issue, but has been studied a lot.

Maybe looking at:

L. De Carli, S. M. Hudson, Geometric remarks on the level curves of harmonic functions, Bull. Lond. Math. Soc. 42 (2010), no. 1, 83–95.

L. Flatto, A theorem on level curves of harmonic functions, J. London Math. Soc. (2) 1 (1969) 470–472.

Z. Y. Wen, L. M. Wu, and Y. Zhang, Set of zeros of harmonic functions of two variables, Harmonic analysis, Tianjin, 1988, Lecture Notes in Mathematics 1494 (Springer, Berlin, 1991) 196–203

will help. I do not know the answer to your specific question.

share|cite|improve this answer
Thank you Andreas for the references. This question it was motivated by the main idea of the Steepest Descent Method and I went curious if I could design a holomorphic function where I could not apply this method. – Leandro Nov 26 '10 at 6:37

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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