Self-similar blowup from arbitrary data for supercritical wave maps with additive noise
Irfan Glogić, Martina Hofmanová, Eliseo Luongo
TL;DR
This work proves that additive corotational noise can drive self-similar blowup in energy-supercritical wave maps with positive probability for arbitrary corotational data. The authors combine a Da Prato–Debussche decomposition with a self-similar change of variables and a Lyapunov–Perron fixed-point argument to establish nonlinear stability of the self-similar blowup profile under stochastic forcing. They further show an irreducibility/controllability property of the stochastic forcing, ensuring the solution can reach neighborhoods in state space with positive probability, which together imply probabilistic blowup via the explicit self-similar profile ${\mathcal U}_T$. The results highlight that noise can reinforce certain singular behaviors in nonlinear wave dynamics, contributing to the broader discourse on regularization-by-noise and stochastic instability in supercritical PDEs, with precise framework built in radial Sobolev spaces and corotational symmetry.
Abstract
We consider stochastically perturbed wave maps from $\mathbb{R}^{1+d}$ into $\mathbb{S}^d$, in all energy-supercritical dimensions $d \geq 3$. We show that corotational non-degenerate Gaussian additive noise leads to self-similar blowup with positive probability for any corotational initial data. The same result without noise is conjectured, but unknown, for large data.
