2
$\begingroup$

Suppose that $\Omega \subseteq \mathbb{R}^n$ is a bounded domain and $u : \Omega \to \mathbb{C}$ solves $-\Delta u = \lambda u$ with Dirichet or Neumann boundary conditions.

Can we say anything about the integrability of $u^{-\alpha}$ for some $\alpha > 0$?

I have seen that in the more general case of a Riemannian manifold we can have points with vanishing order of the order $\sqrt{\lambda}$. Can we hope to do better specialising to the case of a flat manifold with boundary as in the case above?

$\endgroup$
1
  • $\begingroup$ The counterexamples are essentially local (well, they have a global component where you estimate $N^2 \leq C \lambda$, $N$ being the frequency, but that part is optimal as well even for domains, you can construct examples by separating variables), so you shouldn't expect to do better on domains. Whether or not you can actually get that level of integrability is unclear to me, though possibly you can using quantitative (Naber-Valtorta type) singular set estimates. I've not seen this in the literature, would suggest checking out or asking Lagunov/Malinnikova if you haven't already. $\endgroup$
    – user378654
    Commented Jul 28, 2023 at 23:52

0

You must log in to answer this question.

Browse other questions tagged .