Wellposedness of inviscid SQG in the half-plane

In-Jee Jeong, Junha Kim, Hideyuki Miura · 2025

Read Wellposedness of inviscid SQG in the half-plane by In-Jee Jeong, Junha Kim, Hideyuki Miura (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-06601.txt.

Introduction

In-Jee Jeong, Junha Kim, and Hideyuki Miura study a fluid equation on a half-plane, the surface quasigeostrophic equation with no dissipation and a fixed boundary condition. They prove the equation is locally well-posed: starting data in the stated smoothness spaces give a unique solution that behaves well for a short time. The spaces involved allow fairly rough data, covering a wide range of exponents.

They complete the picture with a sharpness result. For typical smooth starting data with compact support, the unique solution does not belong to a slightly stronger space with the exponent pushed to infinity. That limit case is a genuine boundary of the theory. The paper thus maps both where the equation can be solved and where the solution theory stops improving.

Catalog record

CREATORIn-Jee Jeong, Junha Kim, Hideyuki Miura
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE74 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-06601

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Wellposedness of inviscid SQG in the half-plane In-Jee Jeong, Junha Kim, Hideyuki Miura Abstract . We consider the SQG equation without dissipation on the half-plane with Dirichlet boundary condition, and prove local wellposedness in the spaces W 3 , p W^{3,p} and C 2 , β C^{2,\beta} for any 1 < p < ∞ 1<p<\infty and 0 < β < 1 0<\beta<1 . We complement this wellposedness by showing that for generic C 0 ∞ C^{\infty}_{0} initial data, the unique corresponding solution does not belong to W 3 , ∞ W^{3,\infty} . † † footnotetext: Key words: wellposedness, half-plane, surface quasi-geostrophic equation, fluid dynamics 2020 AMS Mathematics Subject Classification: 76B47, 35Q35 1. Introduction 1.1. Inviscid SQG in the half-plane We consider the Cauchy problem for the inviscid surface quasi-geostrophic (SQG) equation in the upper half-plane \mathbb{R}^{2}_{+}=\{(x_{1},x_{2})\in\mathbb{R}^{2};x_{2}>0\} : (SQG) \partial_{t}\theta+u\cdot\nabla\theta=0, where the velocity is given by (BS) \begin{split}u(t,x)=P.V.\int_{\mathbb{R}^{2}_{+}}\left[\frac{(x-y)^{\perp}}{|x-y|^{3}}-\frac{(x-\bar{y})^{\perp}}{|x-\bar{y}|^{3}}\right]\theta(t,y)\,\mathrm{d}y,\end{split} with \bar{y}:=(y_{1},-y_{2}) for y=(y_{1},y_{2}) . That is, up to a multiplicative constant that we neglect, u=R^{\perp}\bar{\theta} , where R^{\perp}=(-R_{2},R_{1}) with R_{i}=\partial_{i}(-\Delta)^{-1/2} and \bar{\theta} is the odd extension of \theta to \mathbb{R}^{2} . As we shall explain below, to study strong solutions for ( SQG ) in the half-plane, it is essential to impose the Dirichlet boundary condition (DBC) \begin{split}\theta(t,x_{1},0)\equiv 0,\end{split} which will be assumed throughout the paper. Since the velocity satisfies the slip boundary conditions, the condition ( DBC ) propagates in time. While dissipative versions of ( SQG ) have been studied quite extensively in domains with boundary ( [ 6 , 5 , 8 , 9 , 29 , 20 , 21 , 22 ] ), culminating in the global wellposedness result with critical dissipation ( [ 7 ] ), for the inviscid SQG, the only wellposedness for strong solutions that we are aware of is the one in Constantin–Nguyen [ 8 ] which gives local wellposedness in H^{1}_{0}\cap W^{2,p} for any 2<p<\infty . While the proof in [ 8 ] is given for any planar bounded domain with smooth boundary, it can be easily adapted to the half-plane case. Given this H^{1}_{0}\cap W^{2,p} wellposedness result, it is natural to ask whether higher Sobolev or Hölder regularity of the initial data can propagate in time for this H^{1}_{0}\cap W^{2,p} solution. More precisely, we investigate what are the smallest Sobolev and Hölder spaces in which ( SQG ) is locally wellposed with ( DBC ). Furthermore, we study the maximal regularity of the unique local-in-time strong solution, for a generic initial datum which is compactly supported and C^{\infty} up to the boundary. 1.2. Main Results Our first result gives local wellposedness of data belonging to either C^{2,\beta} ( 0<\beta<1 ) or W^{3,p} ( 1<p<\infty ), under the Dirichlet boundary condition. Both of these spaces embed into W^{2,q} with some q>2 , and therefore it is a natural extension of the inviscid wellposedness result of [ 8 ] . Before stating the result, let us introduce a simplifying notation: if X is a Banach space of functions on \overline{\mathbb{R}^{2}_{+}} which embeds in C(\overline{\mathbb{R}^{2}_{+}}) , we write X_{0} to be the subspace of X which are compactly supported and vanishes on the boundary of the upper half-plane. (We are using the notation \overline{\mathbb{R}^{2}_{+}} to emphasize that the regularity holds uniformly up to the boundary.) Theorem A . We have local wellposedness for ( SQG ) on \mathbb{R}^{2}_{+} with ( DBC ) in C^{2,\beta}_{0}(\overline{\mathbb{R}^{2}_{+}}) for any 0<\beta<1 and also in W^{3,p}_{0}(\overline{\mathbb{R}^{2}_{+}}) for any 1<p<\infty . Namely, for any initial data \theta_{0}\in C^{2,\beta}_{0}(\overline{\mathbb{R}^{2}_{+}}) ( W^{3,p}_{0}(\overline{\mathbb{R}^{2}_{+}}) , resp.), there exist \delta>0 and a unique corresponding solution satisfying \theta\in L^{\infty}([0,\delta];C^{2,\beta}_{0}(\overline{\mathbb{R}^{2}_{+}})) ( {C}([0,\delta];W^{3,p}_{0}(\overline{\mathbb{R}^{2}_{+}})) , resp.). Remark 1.1 . A few remarks are in order. • As in the \mathbb{R}^{2} case, \begin{split}\int_{0}^{T}\|\nabla\theta(t,\cdot)\|_{L^{\infty}(\mathbb{R}^{2}_{+})}dt<\infty\end{split} controls singularity formation at t=T . Furthermore, the uniqueness statement requires just \theta\in L^{\infty}([0,\delta];C^{1}(\overline{\mathbb{R}^{2}_{+}})) (the proofs in [ 1 , 24 ] for \mathbb{R}^{2} can be easily adapted to the half-plane case). • As a direct consequence of this wellposedness result, C^{\infty}_{0} initial datum has the corresponding unique solution belonging to to C^{2,\beta}\cap W^{3,p} for any \beta<1 and p<\infty during its lifespan. The next result complements the first main result by showing that there is nonexistence of solutions in W^{3,\infty}_{0}(\overline{\mathbb{R}^{2}_{+}}) (and hence also in C^{3}_{0}(\overline{\mathbb{R}^{2}_{+}}) ), even if the initial data is infinitely differentiable up to the boundary. Theorem B . There exist initial data in \theta_{0}\in C^{\infty}_{0}(\overline{\mathbb{R}^{2}_{+}}) such that the corresponding unique solution to ( SQG ) given in Theorem A does not belong to L^{\infty}([0,\delta];W^{3,\infty}(\overline{\mathbb{R}^{2}_{+}})) for any \delta>0 . A sufficient condition ensuring this is to have a point x^{*}\in\partial\mathbb{R}^{2}_{+} such that \partial_{12}\theta_{0}(x^{*}) and \partial_{22}\theta_{0}(x^{*}) are both nonzero. This statement needs to be carefully distinguished with strong illposedness results in C^{k} and W^{k,\infty} for integer k proved in [ 12 , 24 ] for domains without boundary: in that case, C^{\infty} initial data lead to C^{\infty} local in time solutions, and C^{k},W^{k,\infty} illposedness is purely coming from the initial data being exactly C^{k} or W^{k,\infty} and not better. On the other hand, nonexistence in Theorem B is for C^{\infty} initial data and is really coming from the boundary of the domain. Remark 1.2 . Let us clarify that the Dirichlet boundary condition is essential for local wellposedness of strong solutions, at least when the domain is \mathbb{R}^{2}_{+} : it was shown in [ 23 ] that for any compactly supported smooth datum \theta_{0} on \mathbb{R}^{2}_{+} which does not satisfy ( DBC ), there is nonexistence of corresponding L^{\infty}([0,\delta];C^{1}(\overline{\mathbb{R}^{2}_{+}})) solution to ( SQG ) with any \delta>0 . This is an optimal nonexistence result, since one cannot expect wellposedness of strong solutions strictly below C^{1} , this space being scaling critical. 1.3. Discussion on singularity problem for SQG Besides extending the results of [ 8 ] , another motivation of the current work is to explore settings in which finite time singularity formation for ( SQG ) can be established. Singularity formation for ( SQG ) is a notoriously difficult open problem, and there does not seem to be convincing numerical evidence towards blowup for smooth data in domains without boundaries. We refer to [ 11 , 10 , 3 , 4 , 14 , 16 , 18 , 19 ] for some results, against singularity formation, regarding the SQG equation. The problem is still challenging for the generalized models of the form (1.1) \left\{\begin{aligned} &\partial_{t}\theta+u\cdot\nabla\theta=0,\\ &u=\nabla^{\perp}(-\Delta)^{-1+\frac{\alpha}{2}}\theta,\end{aligned}\right. which are sometimes referred to as the \alpha -SQG equation. With this notation, \alpha=1 corresponds to ( SQG ) and \alpha=0 to the vorticity formulation for two-dimensional incompressible Euler equations. A larger value of \alpha makes the kernel for the velocity more singular. Global regularity of smooth solutions for any \alpha>0 in \mathbb{R}^{2} is a challenging open problem. Towards the blowup problem, the half-plane provides a nice stage, the main point being that “patch type” data attached to the boundary cannot be detached in finite time, unless a singularity of some form occurs. Essentially, this is the common setup for all the existing blowup results for ( 1.1 ): [ 26 , 27 , 15 , 28 ] deal with patch solutions attached to the boundary and the very recent work [ 30 ] concerns solutions in properly weighted Hölder spaces, non-vanishing on the boundary. Unfortunately, all of these results

Read the complete text