Breakdown of smooth solutions to the subcritical EPDiff equation

Martin Bauer, Stephen C. Preston, Justin Valletta · 2025

Read Breakdown of smooth solutions to the subcritical EPDiff equation by Martin Bauer, Stephen C. Preston, Justin Valletta (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-2503-19780.txt.

Introduction

Martin Bauer, Stephen Preston, and Justin Valletta wrote a 2025 math paper on when smooth solutions of the EPDiff equation break down. This equation models flows where the velocity field is built from a momentum field through a smoothing operator. It appears in shape analysis and fluid mechanics.

The authors study the case where the smoothing uses a whole number power of the Laplacian. For radial starting data with a sign condition on momentum, they prove the solution's slope blows up in finite time across a wide range of powers. Their method recasts the equation as an ordinary differential equation in Lagrangian coordinates and compares it with a classic equation. The result is a step toward a full conjecture on which powers give solutions that last for all time.

Catalog record

CREATORMartin Bauer, Stephen C. Preston, Justin Valletta
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE49 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-19780

Browse the full library · About the Tower of Babel

Full text

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

Breakdown of smooth solutions to the subcritical EPDiff equation Martin Bauer, Stephen C. Preston, Justin Valletta Abstract We consider the EPDiff equation on ℝ n \mathbb{R}^{n} with the integer-order homogeneous Sobolev inertia operator A = ( − Δ ) k A=(-\Delta)^{k} . We prove that for arbitrary radial initial data and a sign condition on the initial momentum, the corresponding radial velocity solution has C 1 C^{1} norm that blows up in finite time whenever 0 ≤ k < n / 2 + 1 . 0\leq k<n/2+1. Our approach is to use Lagrangian coordinates to formulate EPDiff as an ODE on a Banach space, enabling us to use a comparison estimate with the Liouville equation. Along the way we derive the Green function in terms of hypergeometric functions and discuss their properties. This is a step toward proving the general conjecture that the EPDiff equation is globally well-posed for any Sobolev inertia operator of any real order k k if and only if k ≥ n / 2 + 1 k\geq n/2+1 . Introduction The EPDiff equation: We are interested in a family of non-linear PDEs, referred to as the EPDiff equation, which are given by \Omega_{t}+\nabla_{U}\Omega+(\nabla U)^{T}\Omega+\operatorname{div}(U)\Omega=0,\qquad\Omega=AU, (1) where U:[0,T)\times\mathbb{R}^{n}\to\mathbb{R}^{n} is a time-dependent vector field, A is an L^{2} symmetric, continuous linear operator, called the inertia operator , and \Omega is often regarded as the momentum. This family of equations first arose in the context of the Camassa-Holm equation [ 14 ] , which corresponds to the order-one Sobolev inertia operator A=1-\Delta ; see [ 31 , 32 , 40 ] . More generally, it encompasses many classic one-dimensional fluid models, such as Burgers’, the modified Constantin-Lax-Majda [ 19 , 43 ] , and the Hunter-Saxton equation [ 33 ] . It can thereby be viewed as an n -dimensional generalization of these fluid equations. Moreover, EPDiff admits a geometric interpretation as an Euler-Arnold equation [ 3 ] : it can be realized as the geodesic equation of a right-invariant Riemannian metric on the group of diffeomorphisms; such equations are also called Euler-Poincaré equations, which is how EPDiff acquired its name. The interest in the EPDiff equations is thus further driven by the fundamental role that right-invariant Sobolev metrics on diffeomorphism groups play in template matching and shape analysis [ 20 , 48 , 6 ] , particularly in the LDDMM framework [ 12 ] : in the spirit of Grenander’s pattern theory [ 28 ] this approach represents the differences between shapes as optimal diffeomorphisms between objects, where optimality is measured precisely with respect to a right-invariant metric on the diffeomorphism group. The EPDiff equation thereby arises as the first-order optimality condition. In this context it is also referred to as the template matching equation [ 30 ] . Known results on well-posedness and blowup of solutions: Since its introduction by Holm and Marsden at the turn of the century [ 31 ] , much effort has been dedicated to studying local and global well-posedness of the EPDiff equation. These investigations largely rest on the seminal ideas of Ebin and Marsden [ 22 ] who famously used Arnold’s geometric framework [ 2 ] to prove local well-posedness of the incompressible Euler equations. Using similar methods, the local well-posedness of the EPDiff equation has been established in a quite general setting: assuming that the inertia operator is a pseudo-differential operator of order 2k , local well-posedness has been shown for k\geq\frac{1}{2} , regardless of the dimension n ; see [ 27 , 41 , 24 , 7 , 4 , 46 ] . Under stronger assumptions on the order k , namely when k>n/2+1 , then global well-posedness of EPDiff has been shown; see e.g., the work of Escher, Kolev, Michor, Mumford, and others [ 13 , 23 , 42 , 41 , 7 , 8 , 22 ] . For n=1 the global existence has been extended to k=\frac{3}{2}=\frac{n}{2}+1 , see [ 45 , 10 ] . This raises the question on the existence of breakdown of solutions for EPDiff with an inertia operator of lower order (below the critical index n/2+1 ). Due to the aforementioned connections to one-dimensional fluid models this has been studied in detail in dimension one for the inertia operator A=(\sigma-\Delta)^{k} with \sigma\in\{0,1\} and k\in\{0,1\} : when k=0 , then the one-dimensional EPDiff equation reduces to the inviscid Burgers’ equation, for which breakdown is well-known; for \sigma=0 and k=1/2 this corresponds to the modified Constantin-Lax-Majda for which blowup is known [ 15 , 9 , 45 ] ; for k=1 and \sigma=0 , this corresponds to the Hunter-Saxton equation, for which one can obtain an explicit solution formula [ 34 , 5 ] that leads to a direct proof of breakdown; see also [ 47 ] ; finally, for k=1 and \sigma=1 , this gives the Camassa-Holm equation, for which breakdown is long known [ 14 , 18 ] ; see also the work of McKean [ 38 ] , who obtained the complete picture of the breakdown mechanism for this equation. Combining these blowup results with the previously discussed global existence results, one obtains a complete characterization of global well-posedness (existence of blowup, resp.) for EPDiff with integer-order inertia operator in dimension one. In higher dimensions, the investigation of solution breakdown for the EPDiff equation was only explored much more recently by Chae and Liu [ 16 ] , where they confirmed breakdown for the higher-dimensional Burgers’ equation. Shortly after, breakdown for the higher-dimensional Camassa-Holm equation, corresponding to k=1 , was established by Li, Yu, and Zhai [ 35 ] . In their work they showed that the one-dimensional breakdown mechanism can be adapted to radial solutions in higher dimensions. Here, we emphasize that both of these results have in common, that already the one-dimensional equation admits solutions that break down. In our recent work [ 11 ] we have shown that the EPDiff equation corresponding to A=(\sigma-\Delta)^{2} in dimension n\geq 3 breaks down in finite time, in both the homogeneous (\sigma=0) and the nonhomogeneous (\sigma=1) cases. This was the first instance where the breakdown is a purely higher-dimensional phenomenon, i.e., the one-dimensional equation for k=2 exists globally in time. Summarizing the previous paragraphs, the known cases here are: global well-posedness for any real k>n/2+1 and any n when \sigma>0 and for k=3/2 when n=1 . Solution breakdown has been obtained for any \sigma\geq 0 for k\in\{0,1\} and any n ; for k=2 and n\geq 3 ; and k=\tfrac{1}{2} when n=1 . As a consequence of these results it has been conjectured that the index k=\frac{n}{2}+1 is indeed critical for this property, i.e.: Conjecture 1 . Let A be an elliptic pseudo differential operator of order 2k acting on C^{\infty}(\mathbb{R}^{n},\mathbb{R}^{n}) , that is symmetric and positive w.r.t. the L^{2} -inner product. Then the corresponding EPDiff equation is globally well-posed for all smooth initial data u_{0} if and only if k\geq n/2+1 . Note that the particular case A=(\sigma-\Delta)^{k} with k<\frac{n}{2}+1 in the above conjecture remains open in most cases. The objective of the present paper is to add further evidence for this conjecture by proving new solution breakdown results. Main Contributions: The main result of the present article is the following theorem,

Read the complete text