Inverse Problems for the Return Map in the Class ( ๐’ช C \mathcal{O}_{C} ): Reconstruction and Identifiability

M. El Morsalani ยท 2026

Read Inverse Problems for the Return Map in the Class ( ๐’ช C \mathcal{O}_{C} ): Reconstruction and Identifiability by M. El Morsalani (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-2604-24915.txt.

Introduction

El Morsalani and Barkatou study the inverse problem for the return map on the boundary of a convex core. Splitting return shifts into sideways and normal parts, they derive an exact nonlinear system for thickness, yielding a single-point answer. Linearizing at fixed points introduces an operator Q-star: when Q-star is nonzero, thickness and outer shape are recoverable via global spread. When Q-star is zero, local doubt is settled with nonlinear Ehresmann-type transport plus holonomy maps on closed paths.

Catalog record

CREATORM. El Morsalani
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE70106
LICENSEhttp://creativecommons.org/publicdomain/zero/1.0/
RIGHTSCC0
SOURCEarXiv

Source: https://arxiv.org/abs/2604.24915

Browse the full library ยท About the Tower of Babel

Full text

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

Inverse Problems for the Return Map in the Class ( ๐’ช C \mathcal{O}_{C} ): Reconstruction and Identifiability Thanks: ISTM Laboratory, Chouaib Doukkali University, Morocco. [email protected] M. El Morsalani Abstract We investigate the inverse problem for the return map generated by a geometric round-trip between the boundary of a convex core C C and the boundary of an admissible domain ฮฉ \Omega in the class ๐’ช C \mathcal{O}_{C} . This construction induces a discrete dynamical system on โˆ‚ C \partial C , governed by a scalar thickness function d d that encodes the distance between โˆ‚ C \partial C and โˆ‚ ฮฉ \partial\Omega along outward normal rays. While the forward correspondence from geometry to dynamics is well established, the inverse direction remains largely unexplored. We show that the return map F : โˆ‚ C โ†’ โˆ‚ C F:\partial C\to\partial C uniquely determines the gradient line field of the thickness function and induces a gradient-like dynamical structure. In particular, the critical points of d d , their Morse indices, and the basin decomposition of the dynamics are fully recoverable from the return map. At second order, we prove that the geometric information is not accessed directly, but only through the composite operator A โก ( c ) โ€‹ Hess โˆ‚ C โ€‹ ( d ) โ€‹ ( c ) , A(c)\,\mathrm{Hess}_{\partial C}(d)(c), where the operator A โก ( c ) A(c) is explicitly determined by the geometry of the round-trip mechanism. In particular, at nondegenerate critical points, A โก ( c โˆ— ) = 2 โ€‹ d โ€‹ ( c โˆ— ) โ€‹ ( I โˆ’ d โก ( c โˆ— ) โ€‹ S c โˆ— ) โˆ’ 1 , A(c^{*})=2d(c^{*})\bigl(I-d(c^{*})S_{c^{*}}\bigr)^{-1}, with S c โˆ— S_{c^{*}} denoting the shape operator of โˆ‚ C \partial C . This reveals that the return dynamics encodes the local geometry through a curvature-dependent preconditioning mechanism, combining intrinsic thickness and extrinsic curvature effects. This leads to an intrinsic non-uniqueness of the inverse problem, which we characterize in terms of scaling effects and operator-induced ambiguities at the level of dynamical equivalence. We further show that identifiability up to scaling and dynamical equivalence (i.e., up to reparametrization of trajectories) can be recovered under additional geometric constraints, such as isotropy, alignment of principal directions, or symmetry. Our results establish a fundamental inverse reconstruction principle: the return dynamics encodes the extrinsic geometry of the domain modulo the intrinsic symmetries of the round-trip mechanism, providing a rigorous bridge between geometric analysis, inverse problems, and discrete dynamical systems. Keywords: return map, inverse problem, geometric dynamics, thickness function, second-order reconstruction, operator-induced ambiguity, curvature effects, shape operator, gradient-like systems, Morse theory 2020 Mathematics Subject Classification: Primary 37C25; Secondary 35R30, 37C10, 53C21 1 Introduction Geometric structures often generate nontrivial dynamical mechanisms when interactions between different boundary components are considered. In the present work, we investigate such a phenomenon for a class of domains \mathcal{O}_{C} introduced by Barkatou [ 1 ] , where a convex core C is embedded in a larger domain \Omega\subset\mathbb{R}^{N} . A fundamental feature of this class is that the geometry of \Omega can be encoded by a scalar function defined on the boundary \partial C , namely the thickness function d:\partial C\to\mathbb{R}_{+}, which measures the distance from \partial C to the outer boundary \partial\Omega along outward normal directions. This representation induces a natural geometric mechanism: starting from a point c\in\partial C , one moves along the outward normal direction until reaching \partial\Omega , and then returns to \partial C along the inward normal to \partial\Omega . This round-trip defines a transformation F=\pi\circ\Phi:\partial C\to\partial C, referred to as the return map . Previous works have shown that this map generates a discrete dynamical system on \partial C whose behavior is strongly constrained by the geometry of the domain. In particular: โ€ข the return map admits a first-order expansion revealing a gradient-like structure [ 7 , 8 ] , โ€ข its fixed points coincide with the critical points of the thickness function, โ€ข the induced dynamics is globally convergent and admits a Lyapunov function [ 10 ] , โ€ข the global organization of the dynamics is governed by the topology of \partial C via Morse theory [ 5 , 6 ] . These results establish a correspondence between geometry and dynamics: (\Omega,d)\;\longmapsto\;F. This correspondence can be interpreted as a geometric projection: (\Omega,d)\;\longmapsto\;F, which maps geometric data onto a space of dynamical invariants. From this viewpoint, the return map does not encode the full geometry, but only those features that are invariant under the round-trip mechanism. Importantly, this projection preserves the qualitative orbit structure of the induced dynamics, but not necessarily the precise discrete map. In particular, different geometric configurations may generate return maps with identical orbit structures while differing in their parametrization along trajectories. The inverse problem therefore consists not only in reconstructing the geometry, but in understanding the structure of this projection and the extent to which it can be inverted. Objective of the present work. The purpose of this paper is to investigate the inverse direction of this correspondence. Given a return map F:\partial C\to\partial C, we ask: To what extent does the return dynamics determine the geometry of the domain, and which components of this geometry are fundamentally non-identifiable due to the structure of the round-trip mechanism? More precisely, we study the inverse problem [ 13 , 14 ] F\;\longrightarrow\;(\Omega,d), with a focus on the reconstruction of the thickness function. This problem differs from classical inverse problems in that the data is dynamical rather than boundary-based. In classical inverse problems, the available information typically consists of boundary measurements, such as Dirichlet-to-Neumann maps or response operators, from which one seeks to reconstruct internal geometric or physical properties [ 13 , 14 ] . In contrast, the present work considers a fundamentally different setting, where the observation is not given by boundary data, but by a dynamical operator acting on \partial C . The return map F can thus be interpreted as a dynamical observation operator , encoding geometric information through the induced evolution on the boundary. This shift from boundary measurements to dynamical observations leads to a distinct type of inverse problem, in which the geometry is accessed indirectly through its induced dynamics, rather than through direct boundary responses. Main contributions. We develop a systematic inverse theory for the return map, characterizing both the recoverable geometric information and the intrinsic ambiguities of the reconstruction problem. We show that: โ€ข the return map determines the gradient directions of the thickness function and induces a gradient-like dynamical system whose trajectories follow the same descent lines as the associated gradient flow, up to higher-order geometric corrections; โ€ข the critical points and the qualitative organization of the dynamics are fully recoverable from the return dynamics; โ€ข Explicit second-order structure. At nondegenerate critical points, we compute explicitly the geometric operator governing the linearization: A(c^{*})=2d(c^{*})\bigl(I-d(c^{*})S_{c^{*}}\bigr)^{-1}, where S_{c^{*}} is the shape operator of \partial C . This formula shows that the return dynamics encodes the local geometry through a curvature-dependent preconditioning mechanism, combining intrinsic thickness and extrinsic curvature effects; โ€ข at second order, the geometry is not observed directly, but through the composite operator A(c)\,\mathrm{Hess}_{\partial C}(d)(c), which leads to intrinsic non-uniqueness; โ€ข uniqueness up to scaling and dynamical equivalence

Read the complete text