Local well-posedness of the skew mean curvature flow for large data
Jiaxi Huang, Daniel Tataru
TL;DR
This work proves local well-posedness for the skew mean curvature flow (SMCF) in codimension two with large data in low-regularity Sobolev spaces by reformulating SMCF as a covariant nonlinear Schrödinger system for the second fundamental form $\lambda$ under a dynamic gauge. The authors introduce time-discretization and intrinsic fractional spaces $X^s$, $Y^{s+1}$, and $Z^s$ to handle large data, and develop linearized and parabolic energy estimates to close the argument. Regularized initial data and a passing-to-the-limit strategy yield rough solutions with Hadamard-type well-posedness, including continuous dependence on initial data. The approach avoids nontrapping conditions and low-frequency assumptions of prior work, offering a robust framework for large-data local analysis of geometric quasilinear Schrödinger flows.
Abstract
The skew mean curvature flow is an evolution equation for $d$ dimensional ma\-nifolds embedded in $\mathbb{R}^{d+2}$ (or more generally, in a Riemannian manifold). It can be viewed as a Schrödinger analogue of the mean curvature flow, or alternatively as a quasilinear version of the Schrödinger Map equation. In this article, we prove large data local well-posedness in low-regularity Sobolev spaces for the skew mean curvature flow in dimension $d\geq 2$. This is achieved by introducing several new ideas: (i) a time discretization method to establish the existence of smooth solutions, (ii) constructing the orthonormal frame by a parallel transport method and a lifting criterion, (iii) introducing intrinsic fractional function spaces $X^s\subset H^s$ on a noncompact manifold for any $s>\frac{d}{2}$, such that the $X^s$-norm of the second fundamental form can be propagated well along the quasilinear Schrödinger flow, (iv) deriving a difference equation to prove the uniqueness result for solutions $F\in C^2$, which is independent in the choices of gauge. Our method turns out to be more robust for large data problem.
