Mostly Nonuniformly Sectional Expanding Systems

Basic Information

• Paper ID: 2508.06233
• Title: Mostly nonuniformly sectional expanding systems
• Authors: Vitor Araujo, Luciana Salgado
• Classification: math.DS (Dynamical Systems)
• Publication Date: November 18, 2025 (arXiv v2: November 17, 2025)
• Paper Link: https://arxiv.org/abs/2508.06233

Abstract

This paper introduces the concept of "Mostly Nonuniformly Sectional Expanding" (MNUSE) for singular flows, which unifies and generalizes existing theories including sectional hyperbolicity, asymptotic sectional hyperbolicity, and multi-singular hyperbolicity. The authors construct examples of C¹ nonuniformly sectional hyperbolic sets satisfying MNUSE that are neither sectionally hyperbolic nor asymptotically sectionally hyperbolic. Furthermore, under appropriate smoothness assumptions and conditions on the dimension or Lyapunov exponent signs of the center-unstable subbundle, they prove that attracting MNUSE sets support physical/SRB measures, which are unique when the dynamical system is transitive.

Research Background and Motivation

Problem Context

1. Weak Hyperbolicity Theory for Singular Flows: Since the introduction of singular hyperbolicity by Morales, Pacifico, and Pujals in the late 1990s, researchers have sought to understand the hyperbolic structures of dynamical systems with singularities, such as the Lorenz system.
2. Challenges in Higher-Dimensional Generalizations: In three dimensions, singular hyperbolicity is equivalent to volume expansion; however, in higher-dimensional systems, volume expansion does not imply area expansion (as in Shilnikov-Turaev's "wild strange attractors"), necessitating more refined concepts of sectional hyperbolicity.
3. Limitations of Existing Theories:
   • Sectional Hyperbolicity (SH): Requires uniform area expansion, which is overly restrictive
   • Asymptotic Sectional Hyperbolicity (ASH): Applicable to contracting Lorenz attractors but still limited
   • Multi-Singular Hyperbolicity (MSH): Attempts to characterize star flows but cannot encompass all sectionally hyperbolic systems
   • Strong Nonuniformly Sectional Hyperbolicity (NUSH): Conditions are too strong, unsuitable for certain persistent hyperbolic systems

Research Motivation

Existing definitions of nonuniform hyperbolicity are either too restrictive (such as NUSH) or have limited coverage (such as MSH). This paper aims to propose a broader framework that can:

• Unify various existing hyperbolicity concepts
• Include known important examples (such as Lorenz attractors)
• Guarantee the existence of physical measures
• Provide more flexible analytical tools for singular flows

Core Contributions

1. Introduction of MNUSE Concept: Proposes a new definition of "Mostly Nonuniformly Sectional Expanding" that unifies SH, ASH, and MSH concepts.
2. Clarification of Hierarchical Relationships:
   • Proves that SH, ASH (under two-dimensional center-unstable bundle conditions), and MSH (under specific singularity conditions) are all MNUSE (Theorem A)
   • Constructs counterexamples showing MSH does not contain all SH systems (Theorem B)
   • Constructs examples of MNUSE that are neither SH, ASH, nor MSH (Theorem C)
3. Physical Measure Existence Theorem: Proves that under appropriate conditions (two-dimensional center-unstable bundle or positive Lyapunov exponent condition), attracting MNUSE sets support a unique ergodic hyperbolic physical/SRB measure (Theorem D).
4. Important Counterexamples:
   • Exhibits sectionally hyperbolic attracting sets containing non-Lorenz-type singularities
   • Constructs sectionally hyperbolic attracting sets without Lorenz-type singularities
   • Provides examples of robustly transitive systems without controlled decomposition (Example 2.2)

Detailed Methodology

Problem Formulation

Study weak hyperbolic structures of singular flows (allowing hyperbolic singularities to be accumulated by regular orbits), specifically:

• Input: Flow φᵗ generated by smooth vector field X and its compact invariant set Λ
• Objective: Characterize the expansion properties of the center-unstable subbundle Eᶜᵘ
• Constraints: Allow nonuniformity (validity only on sets of positive measure)

Core Definition System

1. Partially Hyperbolic Decomposition

A compact invariant set Λ is partially hyperbolic if there exists a continuous DXᵗ-invariant Whitney decomposition: $T_Λ M = E^s \oplus E^{cu}$ satisfying:

• Domination: $\|DX^t|_{E^s_x}\| \cdot \|DX^{-t}|_{E^{cu}_{X^tx}}\| \leq e^{-λt}$
• Uniform Contraction: $\|DX^t|_{E^s_x}\| \leq e^{-λt}$

2. MNUSE Definition (Definition 3.10)

Let Λ be a partially hyperbolic attracting set. If there exist a set Γ ⊂ U of positive measure and η < 0 such that: $\limsup_{T\to\infty} \frac{1}{T}\int_0^T \log\|\wedge^2(Df|_{E^{cu}_{X^sx}})^{-1}\|ds \leq η, \quad x\in Γ$ where f = Xᵗ is the discrete-time map.

Physical Meaning: Requires that any 2-plane in the center-unstable direction undergoes exponential expansion on average over a set of positive measure.

3. NUSE Definition (Stronger Version)

If there exist a set Γ of positive measure and η < 0 such that: $\limsup_{n\to\infty} \frac{1}{n}\sum_{i=0}^{n-1} \log\|(P^1|_{N^{cu}_{f^ix}})^{-1}\| \leq η$ where $N^{cu}_x = E^{cu}_x \cap G(x)^\perp$ is the projection of the center-unstable direction onto the orthogonal complement of the flow direction.

Technical Innovations

1. Weakening of Uniformity Requirements

Key differences from classical definitions:

• SH: Requires uniform expansion for all points and all times
• MNUSE: Requires expansion only on positive measure sets in average sense
• Allows individual trajectories or time intervals to violate expansion conditions

2. Measure-Theoretic Perspective

Exploiting the existence of physical measures:

• For SH attracting sets, finitely many physical measures μ₁,...,μₚ are known to exist, covering almost all points
• Via Birkhoff's ergodic theorem, points in the basin B(μᵢ) satisfy MNUSE conditions
• Key estimate (Formula 8): $\lim_{t\to\infty}\frac{1}{t}\int_0^t \log\|\wedge^2(DX^T|_{E^{cu}_{X^sx}})^{-1}\|ds = \int_Λ \log\|\wedge^2(DX^T|_{E^{cu}_x})^{-1}\|dμ \leq \log\frac{e^{-λT}}{K}$

3. Linear Poincaré Flow Technique

For handling MSH, the linear Poincaré flow is introduced: $P^t_x: N_x \to N_{X^tx}, \quad P^t_x = O_{X^tx} \circ DX^t(x)$ where $N_x = \{v\in T_xM: \langle v, X(x)\rangle = 0\}$ is the orthogonal complement of the flow direction.

Experimental Setup

Theoretical Verification Strategy

This is a pure theoretical mathematics paper employing constructive proofs and counterexamples:

1. Proof Strategy for Theorem A:
   • Utilize known physical measure existence results
   • Establish time average equals space average via Birkhoff's ergodic theorem
   • Verify MNUSE conditions hold on measure basins
2. Counterexample Construction for Theorem B:
   • Example 5.1: Extension of classical Lorenz attractor including non-Lorenz-type saddle points σ₁, σ₂ (with complex expanding eigenvalues)
   • Example 5.2: Construction via Dehn surgery of sectionally hyperbolic attracting set without Lorenz-type singularities
3. Counterexample Construction for Theorem C:
   • Example 6.1 (corresponding to conclusion 1): Geometric Lorenz construction based on Manneville map, containing non-hyperbolic periodic orbits O(p±), measure-preserving but non-uniformly expanding
   • Example 6.2 (corresponding to conclusion 2): Bruin-Farias geometric Lorenz attractor with neutral singularity, neither SH nor ASH

Key Technical Tools

1. Kingman Subadditive Ergodic Theorem: Handles long-time averages of non-additive quantities (such as logarithms of determinants)
2. Controlled Decomposition Theory: Utilizes results by Araujo-Arbieto-Salgado on controlled decompositions for singular flows
3. Pesin Theory: Constructs unstable manifolds for nonuniformly hyperbolic measures

Experimental Results

Main Results

Theorem A (Inclusion Relations)

Conclusion: The following attracting sets are all MNUSE:

1. Sectionally hyperbolic (SH) attracting sets
2. Asymptotically sectionally hyperbolic (ASH) attracting sets with two-dimensional center-unstable bundle (saddle-type singularities)
3. Multi-singular hyperbolic (MSH) attracting sets (singularities are active and have the same index)

Proof Highlights:

• For SH: Utilize existence of finitely many physical measures μ₁,...,μₚ, verify MNUSE on their basin union Γ = B(μ₁) ∪ ... ∪ B(μₚ)
• For ASH: Key is proving physical measures do not support singularities (otherwise entropy would be zero but Lyapunov exponents positive, contradiction)
• For MSH: Utilize results by Crovisier et al., showing MSH (under given conditions) implies SH

Theorem B (Separation)

Conclusion: There exist star vector fields in Cʳ open sets whose sectionally hyperbolic attracting sets are not MSH.

Construction Method:

• Example 5.1: Lorenz attractor A = Λ ∪ Wᵘ(σ₁) ∪ Wᵘ(σ₂), where σ₁, σ₂ have complex expanding eigenvalues, not satisfying Lorenz-type conditions
• Example 5.2: Glue Anosov flow and vector field with saddle point via Dehn surgery, obtaining SH attracting set without Lorenz-type singularities

Significance: Proves MSH concept cannot encompass all SH systems, clarifying theoretical hierarchy.

Theorem C (Proper Inclusion)

Conclusion: There exist MNUSE attracting sets that are:

1. ASH but neither SH nor MSH
2. Neither SH, nor ASH, nor MSH

Construction:

• Example 6.1 (corresponding to conclusion 1): Geometric Lorenz construction based on Manneville map, containing non-hyperbolic periodic orbits O(p±), measure-preserving but non-uniformly expanding
• Example 6.2 (corresponding to conclusion 2): Bruin-Farias neutral singularity Lorenz attractor, neither SH nor ASH

Verification: Both examples verified by proving existence of unique physical/SRB measure and validity of entropy formula, confirming MNUSE.

Theorem D (Physical Measure Existence)

Conclusion: Let K be a forward-invariant compact set of a C² vector field satisfying MNUSE. If one of the following holds:

1. dim Eᶜᵘ = 2, or
2. No negative Lyapunov exponents on Eᶜᵘ: $\liminf_{T\to\infty}\frac{1}{T}\log\|\wedge^2(Dφ^T(x)v)^{-1}\|\geq 0$

then the attracting set Λ supports an ergodic hyperbolic physical/SRB measure, which is unique when transitive.

Proof Strategy:

1. Utilize Theorem 7.1: MNUSE condition (5) implies NUSE condition (6) on positive measure sets
2. Apply results from 8, Corollaries D & G: Partial hyperbolicity + NUSE ⇒ physical/SRB measure existence

Important Findings

1. Controlled Decomposition Non-Necessity (Example 2.2)

Constructs suspension flow on 4-torus, demonstrating:

• Existence of robustly transitive set with physical measure satisfying NUSE
• But no controlled decomposition exists (expanding directions of periodic points lie in "stable" subbundle)

Significance: Shows MNUSE framework is more flexible than traditional partially hyperbolic theory.

2. Generalization of Hyperbolic Lemma (Lemma 3.4)

All compact invariant subsets without singularities in SH, ASH, MSH sets are uniformly hyperbolic. This bridges nonuniform and uniform theories.

3. Sufficiency of Measure-Theoretic Conditions

Key insight: For smooth vector fields, verifying expansion on positive measure sets combined with ergodic theory suffices to derive global dynamical properties.

Related Work

Development of Singular Hyperbolicity Theory

1. Singular Hyperbolicity (1999): Morales-Pacifico-Pujals introduced for understanding Lorenz systems, requiring volume expansion of center-unstable bundle.
2. Sectional Hyperbolicity (2008): Metzger-Morales generalized to higher dimensions, requiring area expansion of arbitrary 2-planes.
3. p-Sectional Hyperbolicity (2019): Salgado introduced intermediate-dimensional expansion concept.
4. Multi-Singular Hyperbolicity (2021): Bonatti-da Luz introduced for characterizing star flows; Crovisier et al. (2020) provided equivalent definitions.
5. Asymptotic Sectional Hyperbolicity (2017): Morales-San Martin introduced for handling Rovella attractors, requiring average-sense expansion.
6. Strong Nonuniformly Sectional Hyperbolicity (2011): Arbieto-Salgado introduced, but conditions too strong.

Positioning of This Work

MNUSE as the weakest nonuniform expansion condition:

• Upward Compatible: Contains SH, ASH, MSH as special cases
• Downward Guarantees: Still implies physical measure existence
• Gap Filling: Covers examples previously inaccessible to existing theories

Physical Measure Theory

• SH Case: Araujo-Pacifico-Pujals-Viana (2009) proved finiteness
• ASH Case: Araujo-Salgado-Sousa (2025) proved existence
• This Work's Contribution: Unified framework for existence and uniqueness

Conclusions and Discussion

Main Conclusions

1. Theoretical Unification: MNUSE concept successfully unifies multiple weak hyperbolicity theories for singular flows, establishing clear hierarchical structure: $\text{SH, ASH, MSH} \subsetneq \text{MNUSE}$
2. Physical Measure Guarantee: Under mild conditions (two-dimensional center-unstable or positive Lyapunov exponent), MNUSE attracting sets support physical/SRB measures.
3. Importance of Counterexamples: Through carefully constructed counterexamples, clarifies true inclusion relationships between concepts, proving limitations of existing theories.

Limitations

1. Smoothness Requirements: Theorem D requires C² smoothness, necessary for physical measure construction but limiting applicability.
2. Dimensional Restrictions: Results on asymptotic sectional hyperbolicity apply only to two-dimensional center-unstable bundles; higher-dimensional cases remain open.
3. Role of Controlled Decomposition: Although Example 2.2 shows controlled decomposition non-essential, proofs still heavily use it; its precise role remains unclear.
4. Singularity Conditions: Theorem A imposes specific requirements on singularities for ASH and MSH (saddle-type, active, same index); general cases unresolved.

Future Directions

The paper explicitly proposes two research directions:

1. Relaxing Continuity Requirements:
   • Hölder Continuity: Citing Mi-Cao-Yang (2022), Hölder continuous decompositions may replace controlled decompositions
   • Mere Continuity: Citing Cao-Mi-Zou (2025), possibly only continuous decompositions needed for constructing Pesin unstable manifolds
2. Generalization to Broader Settings:
   • Study higher-dimensional center-unstable bundles
   • Explore weaker Lyapunov exponent conditions
   • Investigate non-attracting cases (e.g., saddle sets)
3. Implicit Directions (reviewer conjecture):
   • Statistical properties: mixing rates, central limit theorems
   • Topological properties: topological entropy, periodic orbit growth rates
   • Stability: structural stability, statistical stability

In-Depth Evaluation

Strengths

1. Theoretical Depth

• Conceptual Innovation: MNUSE definition cleverly balances generality and usability, encompassing known examples while guaranteeing physical measure existence
• Clear Hierarchy: Systematic theorems and counterexamples completely characterize relationships between hyperbolicity concepts
• Technical Sophistication: Synthesizes ergodic theory, differential topology, and measure theory

2. Artistry of Example Construction

• Example 5.1: Cleverly exploits Lorenz system geometry, adding non-Lorenz-type saddle points
• Example 6.1: Elevates fine properties of one-dimensional dynamics (Manneville map) to flows
• Example 2.2: Suspension flow of Tahzibi torus diffeomorphism, demonstrating controlled decomposition non-necessity

3. Writing Quality

• Clear structure: Introduction-definitions-theorems-examples-proofs with rigorous logic
• Sufficient motivation: Each definition has clear physical or mathematical background
• Complete technical details: Key proof steps adequately explained or cited

4. Practical Value

• Provides unified framework for studying concrete systems (e.g., Lorenz families)
• Physical measure existence theorem has broad application prospects
• Counterexamples crucial for understanding theory boundaries

Weaknesses

1. High Technical Threshold

• Requires deep background in differential dynamical systems
• Multiple references to authors' previous work; difficult for non-specialists
• Some proof details (e.g., Theorem D) depend on external literature

2. Missing Computational Verification

• All examples are theoretical constructions lacking numerical simulation verification
• Specific numerical characteristics of physical measures in Example 6.1 not provided
• Unclear how to verify MNUSE conditions in practical systems

3. Insufficient Analysis of Condition Necessity

• Are the two conditions in Theorem D (two-dimensional or positive exponent) optimal?
• Can C² smoothness requirement be lowered?
• Is partial hyperbolicity necessary?

4. Limited Connection to Physical Applications

• No discussion of MNUSE applications in physical systems (fluids, climate models, etc.)
• Physical meaning of physical measures insufficiently elaborated
• Connections to turbulence theory not addressed

Impact Assessment

Contributions to the Field

• Theory Refinement: Fills important gap in nonuniform hyperbolicity theory
• Methodology: Establishes paradigm of measure-theoretic perspective on nonuniformity
• Problem Clarification: Counterexamples precisely delineate theory boundaries

Potential Applications

1. Dynamical System Classification: Provides finer classification tools for singular flows
2. Physical System Analysis: Possible applications to Navier-Stokes equations, climate models, etc.
3. Algorithm Design: MNUSE verification may lead to new numerical algorithms

Reproducibility

• Theoretical Results: Complete proofs, strong reproducibility
• Example Construction: Clear principles but requires specialized knowledge
• Numerical Verification: Lacks code or specific parameters, difficult to reproduce

Applicable Scenarios

Theoretical Research

• Study ergodic properties of singular flows
• Construct new hyperbolic or partially hyperbolic examples
• Develop nonuniform hyperbolicity theory

Practical Applications

• Analyze fluid systems with singularities
• Study long-term behavior of climate models
• Understand dynamics of chemical reaction networks

Limiting Scenarios

• Discrete dynamical systems (require appropriate modifications)
• Conservative systems (no attracting sets)
• High-dimensional systems (some results limited to low dimensions)

Key References

1. 25 Morales-Pacifico-Pujals (1999): Foundational work on singular hyperbolicity
2. 21 Metzger-Morales (2008): Definition of sectional hyperbolicity
3. 24 Morales-San Martin (2017): Asymptotic sectional hyperbolicity
4. 10,14 Bonatti-da Luz (2021), Crovisier et al (2020): Multi-singular hyperbolicity
5. 9 Arbieto-Salgado (2011): Strong nonuniformly sectional hyperbolicity
6. 8 Araujo-Salgado-Sousa (2025): Direct predecessor of this work, physical measure existence
7. 5 Araujo-Pacifico (2010): Comprehensive reference on three-dimensional flows
8. 18 Ledrappier-Young (1985): Entropy formula for SRB measures

Overall Assessment: This is a high-quality theoretical mathematics paper making important contributions to nonuniform hyperbolicity theory for singular flows. By introducing the MNUSE concept, it successfully unifies multiple existing theories and clarifies theoretical boundaries through carefully constructed examples. The main theorems, particularly Theorem D, provide a general framework for physical measure existence. The paper demonstrates high technical depth and theoretical completeness with significant impact on dynamical systems. Main limitations include lack of numerical verification and practical application discussion, with optimality of certain technical conditions remaining open.