Tensor tomography and frame flow ergodicity for magnetic flows in higher dimensions

Louis-Brahim Beaufort · 2026

Read Tensor tomography and frame flow ergodicity for magnetic flows in higher dimensions by Louis-Brahim Beaufort (2026) 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-2604-12495.txt.

Introduction

Louis-Brahim Beaufort extends geodesic flow results to flows with magnets. First, he studies the magnetic ray transform and proves a tensor tomography result. Second, he treats ergodicity of the magnetic frame flow under pinching. Prior work guides the approach. New Pestov identities power the proofs. They extend and improve prior work. The framework adapts covariant differentiation, torsion, curvature, and Jacobi fields to magnets. His tensor extends a proposed sectional curvature.

Catalog record

CREATORLouis-Brahim Beaufort
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE134167
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2604.12495

Browse the full library · About the Tower of Babel

Full text

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

Tensor tomography and frame flow ergodicity for magnetic flows in higher dimensions Louis-Brahim Beaufort Abstract . We extend two results from the theory of geodesic flows to the magnetic setting on manifolds of arbitrary dimension. First, we investigate the magnetic ray transform and establish a tensor tomography result. Second, we define and analyze the ergodicity of the magnetic frame flow under a pinching condition, building on work of Cekić-Lefeuvre-Moroianu-Semmelmann. These generalizations rely on new Pestov identities tailored to the magnetic flow, which extend and improve identities derived by Dairbekov-Paternain. In the process, we develop a framework that adapts several concepts of Riemannian geometry to the magnetic context, including covariant differentiation, torsion, curvature, and Jacobi fields. Notably, our curvature tensor generalizes the magnetic sectional curvature recently proposed by Assenza. 2020 Mathematics Subject Classification 37D30, 37D40, 53C15 1. Introduction 1.1. Background The goal of this article is to generalize several results on geodesic flows over closed, negatively curved Riemannian manifolds to the magnetic setting. Given a closed Riemannian manifold (M,g) and a differential 2 -form \mg on M (which is customarily closed), a magnetic geodesic \gamma is a smooth curve \gamma:I\to M , where I\subset\mathbb{R} is an interval, which satisfies the ordinary differential equation \displaystyle\nabla_{\gamma^{\prime}}^{LC}\gamma^{\prime}=\Omega\gamma^{\prime}, where \nabla^{LC} denotes the Levi-Civita connection associated with (M,g) and by a slight abuse of notation \Omega denotes the endomorphism obtained from the 2 -form \mg by metric duality. More precisely, if p\in M , x,y\in T_{p}M , we have \mg(x,y)=\langle\Omega x,y\rangle . The standard theory of ODE’s shows that for every initial data (p,v)\in TM , there is a unique magnetic geodesic \gamma defined over \mathbb{R} such that \gamma(0)=p , \gamma^{\prime}(0)=v . The magnetic flow \varphi_{t}^{\mg} , acting on the tangent bundle TM , is then defined by \varphi_{t}^{\mg}(p,v)=(\gamma(t),\gamma^{\prime}(t)) . In particular, when \mg=0 , \varphi_{t}^{0} is the geodesic flow. Geodesic flows and magnetic flows share several important properties. For instance, since magnetic trajectories have constant speed, they induce flows on the sphere bundles S^{r}M=\{(p,v)\in TM,\|v\|=r\} . Both flows preserve the standard volume (the Liouville volume form) on these bundles. Moreover, when \mg is closed, both flows are Hamiltonian. There are still notable differences. Indeed, magnetic flows are in general not Reeb flows for any contact form. Also, nontrivial reparametrizations of magnetic geodesics are not magnetic geodesics, which implies that there is no affine connection on M such that magnetic geodesics correspond to the geodesics for this connection. Nevertheless, one can still study magnetic flows using the basic geometric framework associated with the Riemannian metric g by introducing magnetic analogues of Riemannian objects, though this requires lengthy computations. Over surfaces, this approach turned out to be fruitful, enabling some authors to define a magnetic Gauss curvature and derive useful integral identities [ MP11 ] . Using these tools, some deep results [ Ass+24 , Ech25 ] were generalized from the Riemannian to the magnetic setting. In higher dimensions, computations become significantly harder, and few results have been obtained until very recently, when a dynamically relevant notion of magnetic curvature appeared [ Ass24 , AMT25 , Ter26 ] along with a notion of magnetic Jacobi fields, see also [ MT26 , BMR26 ] for applications. A relevant notion of curvature is essential for deriving a magnetic variant of a classical Riemannian identity, called the Pestov identity. Introduced by Mukhometov [ Muk75 ] , this identity has played a central role in studying geometric inverse problems and spectral and dynamical rigidity [ GK80 , CS98 , DS03 , PU05 , PSU13 , GL19 ] . More recently, the Pestov identity was instrumental in obtaining sharper pinching estimates in the ergodicity of frame flows [ Cek+24 , Cek+24a ] . One can see this formula as a dynamically motivated variant of the Bochner/Weitzenböck identity [ Cek+25 ] . A Pestov identity was established for magnetic surfaces and used successfully in the study of the magnetic ray transform in dimension 2, see [ Ain13 ] . A natural question is therefore whether magnetic Pestov identities exist in higher dimensions and whether they can be applied to similar problems. 1.2. Magnetic geometry Inspired by the recent derivation of the Pestov identity in [ Cek+25a ] , we develop a magnetic differential calculus by lifting magnetic dynamics from the sphere bundle SM to the orthonormal frame bundle FM . More precisely, we associate a Cartan connection on FM\to SM to a distinguished lift \Phi_{t}^{\mg} of the magnetic flow \varphi_{t}^{\mg} on the unit tangent bundle SM . (Note: no prior knowledge of Cartan geometry is required to read this paper.) This structure represents a deformation of the classical Riemannian geometry induced by the magnetic field. The choice of the lift is nontrivial and can be interpreted in terms of torsion, see Remark 4.2 ; it is also related to the “covariant derivative” \widetilde{\mathcal{D}} and the anisotropic Lorentz force \widetilde{\Omega} introduced in [ AMT25 ] . In particular, we are able to show how this "covariant derivative" operator \widetilde{\mathcal{D}} of [ AMT25 ] , which lacked a geometrical meaning, can be realized as a covariant derivative associated with our connection. We call this special lift \Phi_{t}^{\mg} the magnetic frame flow , and denote its generator \xs . We note that our definition does not coincide with the magnetic frame flow of [ MT26 ] . Using this framework, we geometrically derive notions of curvature, torsion, and Jacobi fields. The resulting curvature tensor \mathcal{R}^{\mg} , computed in Proposition 3.14 , is defined over the unit sphere bundle SM and takes as input tangent vectors to M . Furthermore, \mathcal{R}^{\mg} generalizes the magnetic sectional curvature operator from [ Ass24 ] , which we call the tangential magnetic sectional curvature and denote by \mathcal{M}^{\mg} . More precisely, \displaystyle\mathcal{R}^{\mg}_{(p,v)}(x,v)v=\mathcal{M}^{\mg}_{(p,v)}x,\quad\forall p\in M,\forall v,x\in S_{p}M. Finally, our Jacobi fields correspond to those defined in [ AMT25 ] . Our work thus provides a firm geometric foundation for these concepts. We note that there is a notion of magnetic torsion, which is never zero when \mg\neq 0 (see Lemma 3.13 ). Thus, it is not surprising that the curvature tensor \mathcal{R}^{\mg} lacks symmetries, which is exemplified by a nontrivial Bianchi identity (Proposition 3.15 ). As a result, we encounter unexpected computational difficulties when attempting to generalize several Riemannian arguments to the magnetic case. In the particular case where \mg is a constant multiple of an almost complex structure, most formulas simplify considerably. This is especially true for (constant multiples of) Kähler complex structures, yielding better results; see for instance Remark 3.16 and Theorem 1.3 . We then establish several Pestov identities. First, following [ Cek+25a ] , we prove a universal identity on the frame bundle. The choice of lift \xs is crucial here in order to obtain a simple and useful formula, see Remarks 4.2 and 4.7 . Then, we deduce Pestov identities on tensor bundles over SM constructed from the normal bundle via natural operations. This derivation extends the work of [ Cek+25a ] and generalizes the Pestov identity on vector bundles with connection from [ Gui+16 ] . We also derive a localized Pestov identity for functions over the unit tangent bundle, extending [ Gui+16 ] . Although we limit ourselves to this case, our formalism applies equally

Read the complete text