Explicit bounds on foliated surfaces and the Poincaré problem

Basic Information

• Paper ID: 2511.08388
• Title: Explicit bounds on foliated surfaces and the Poincaré problem
• Author: Stefania Vassiliadis
• Classification: math.AG (Algebraic Geometry)
• Publication Date: November 11, 2025 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2511.08388
• Affiliation: King's College London

Abstract

This paper resolves the Poincaré problem (in the formulation of Cerveau and Lins Neto), providing a linear bound in genus $g$ on the degree of a general leaf of a foliation of general type. The author studies the birational geometry of foliations through the minimal model program (MMP) framework, extending the methods of Spicer-Svaldi and Pereira-Svaldi by investigating the set of pseudo-effective thresholds of adjoint foliations, proving they satisfy the descending chain condition (DCC) and providing explicit universal lower bounds. These results yield effective birational statements for adjoint divisors of the form $K_{\mathcal{F}} + \tau K_X$.

Research Background and Motivation

Core Problem

The research addresses the classical Poincaré problem: Given a foliation $\mathcal{F}$ on $\mathbb{P}^2$, can one bound the degree of algebraic solutions (leaves) in terms of the degree of the foliation and the genus of the solutions? This problem was first posed by Poincaré in 1891 and is central to the theory of algebraically integrable ordinary differential equations.

Significance of the Problem

1. Theoretical significance: Understanding when algebraically integrable ODEs admit algebraic integral solutions is a fundamental question in algebraic geometry
2. Geometric significance: Provides deep understanding of the birational geometry of foliations
3. Applied value: Provides quantitative tools for the classification and study of properties of foliations

Limitations of Existing Methods

1. Pereira (2002): Proved that the degree of a general leaf of a foliation of general type can be bounded, but the bound depends on the degree and genus of the foliation and provides no explicit bound
2. Pereira-Svaldi (2019): For non-equivariant fibrations of genus $g \geq 2$, provides an explicit bound but with exponential dependence on genus $g$
3. Core obstacle: Lack of universal bounds for adjoint foliations $(n,m) \in \mathbb{Z}_{>0} \times \mathbb{Z}_{>0}$ ensuring $h^0(X, K_{\mathcal{F}}^m \otimes K_X^n) \neq 0$

Research Motivation

The core motivation of this paper is to provide the first explicit answer to Problem 1.1, namely to find universal thresholds guaranteeing non-vanishing of adjoint divisors and birationality, thereby improving the exponential bound of Pereira-Svaldi to a linear bound in genus $g$.

Core Contributions

The main contributions of this paper include:

1. DCC property of pseudo-effective thresholds (Theorem 5.5): Proves that the set of pseudo-effective thresholds $R_{2,\eta,I,\epsilon}$ satisfies the descending chain condition, a foundational result for studying the birational geometry of foliations
2. Explicit universal lower bounds (Theorems 5.11 and 5.28):
   • For canonical foliations on smooth surfaces, pseudo-effective thresholds satisfy $\tau(X,\mathcal{F}) \geq \tau_0 := \frac{1}{3(2(2(3(2 \cdot 142151296)! + 1))^{128(3(2\cdot142151296)!+1)^5})!}$
   • Provides explicit constants $M_0(\epsilon)$ ensuring $|m(K_{\mathcal{F}} + \epsilon K_X)|$ defines a birational map
3. Existence of adjoint MMP (Theorem 4.4): For any $\epsilon > 0$, proves existence of $K_{(X,\mathcal{F},\Delta)_\epsilon}$-MMP with controlled singularity variation, extending the result of SS23 for $\epsilon \ll 1$
4. Explicit solution to the Poincaré problem (Theorem 6.1): For foliations on $\mathbb{P}^2$ birationally equivalent to non-equivariant fibrations of genus $g \geq 2$, the degree of a general leaf $F$ satisfies $\deg F \leq M_0(\tau_0)\left(\frac{1}{\tau_0} + 1\right)(4g-4) \cdot \frac{1}{\tau_0} \deg \mathcal{F}$ This is a linear bound in $g$, significantly improving the previous exponential bound
5. Technical innovations:
   • Extends the adjoint foliation framework to arbitrary $\epsilon > 0$
   • Establishes $\epsilon$-adjoint log canonical singularity theory
   • Develops cone and contraction theorems for foliated surfaces

Detailed Methodology

Task Definition

Given a foliated triple $(X, \mathcal{F}, \Delta)$, where:

• $X$ is a normal projective surface
• $\mathcal{F}$ is a rank one foliation
• $\Delta \geq 0$ is a boundary divisor

Core objectives: Study the birational geometry of the adjoint divisor $K_{(X,\mathcal{F},\Delta)_\epsilon} := K_{\mathcal{F}} + \Delta^{n-inv} + \epsilon(K_X + \Delta^{n-inv})$, particularly:

1. Determine the pseudo-effective threshold $\tau(X,\mathcal{F},\Delta) := \sup\{t \in \mathbb{R} | K_{(X,\mathcal{F},\Delta)_t} \text{ is pseudo-effective}\}$
2. Find explicit constants $m$ ensuring $|m(K_{\mathcal{F}} + \epsilon K_X)|$ defines a birational map

Theoretical Framework: Adjoint Foliations

Core idea: Rather than studying the canonical divisor $K_{\mathcal{F}}$ alone, consider divisors of the form $K_{\mathcal{F}} + \epsilon K_X$ ($0 < \epsilon \ll 1$), which allows leveraging classical results on $K_X$.

$\epsilon$-adjoint log canonical divisor (Definition 2.11): $K_{(X,\mathcal{F},\Delta)_\epsilon} := K_{\mathcal{F}} + \Delta^{n-inv} + \epsilon(K_X + \Delta^{n-inv})$

$\epsilon$-adjoint log canonical singularities (Definition 2.12): For any birational morphism $\pi: X' \to X$, $K_{(X',\mathcal{F}',\Delta')_\epsilon} = \pi^* K_{(X,\mathcal{F},\Delta)_\epsilon} + E$ where $E = \sum a_i E_i$ is $\pi$-exceptional with $a_i \geq -(i_{\mathcal{F}}(E_i) + \epsilon)$

Core Technique: Adjoint MMP

Theorem 4.4 (Existence of adjoint MMP): For any $\epsilon > 0$ and foliated triple $(X,\mathcal{F},\Delta)$ satisfying the conditions, there exists a $K_{(X,\mathcal{F},\Delta)_\epsilon}$-MMP: $X \xrightarrow{f_0} X_0 \xrightarrow{f_1} X_1 = Y$

where:

• First phase $f_0$: Contracts all $K_{\mathcal{F}} + \Delta^{n-inv}$-non-positive curves
• Second phase $f_1$: Contracts all curves that are $K_{X_0} + \Delta_0$-negative and $K_{\mathcal{F}_0} + \Delta_0^{n-inv}$-positive

Key properties:

1. If $K_{(X,\mathcal{F},\Delta)_\epsilon}$ is pseudo-effective, then $K_{(Y,\mathcal{F}',\Delta')_\epsilon}$ is nef
2. If $K_{(X,\mathcal{F},\Delta)_\epsilon}$ is not pseudo-effective, then $Y$ admits a fibration-type contraction
3. Singularity control: If $(X,\Delta)$ is $\eta$-lc, then $(X_i, \Delta_i)$ is $\eta' = \frac{\epsilon\eta}{\epsilon+1}$-lc

Technical innovations:

• Extension to arbitrary $\epsilon > 0$ (only $\epsilon \ll 1$ in SS23)
• Decomposition technique via Lemma 3.8, splitting adjoint MMP into two phases
• Singularity control on the base space via Proposition 3.10

Study of Pseudo-effective Thresholds

Definition of pseudo-effective threshold set (5.4.3): $R_{2,\eta,I,\epsilon} := \{\tau(X,\mathcal{F},\Delta) | \dim X = 2, (X,\mathcal{F},\Delta) \text{ satisfy conditions}\}$

Theorem 5.5 (DCC property):

1. If $I \subset [0,1]$ is finite, then $R_{2,\eta,I,\epsilon} \cap (0,\delta)$ is finite for any $\delta > 0$
2. If $I \subset [0,1]$ satisfies DCC, then $R_{2,\eta,I,\epsilon}$ satisfies DCC

Proof strategy:

• Run $K_{(X,\mathcal{F},\Delta)_\tau}$-MMP to obtain $\phi: X \to Y$
• Apply adjunction formula on general fiber $F$: $K_F + \Delta'|_F + \frac{1}{\tau}(K_{\mathcal{G}} + \Delta'^{n-inv})|_F \equiv_{\mathbb{R}} 0$
• Intersect with ample divisor $A$ to show $\frac{1}{\tau}$ belongs to an ACC set

Establishment of Explicit Lower Bounds

Theorem 5.11 (Explicit lower bound for pseudo-effective threshold): For log canonical foliations on $\eta$-lc surfaces, $\tau(X,\mathcal{F},\Delta) \geq \tau'(\eta) := \frac{1}{3 \cdot \left\lfloor 2\left(\frac{2\eta\tau}{\tau+1}\right)^{(2)^7/(\frac{\eta\tau}{\tau+1})^5}\right\rfloor!}$

Proof strategy:

1. Initial test: Consider $K_{\mathcal{F}} + \frac{1}{6}(K_X + \Delta)$
2. MMP analysis: Run $K_{(X,\mathcal{F},\Delta)_{1/6}}$-MMP
3. Case classification (Proposition 5.9):
   • If $K_X + \Delta$ is pseudo-effective, then $\tau = +\infty$
   • If MMP terminates in a fibration over a curve, then $\tau > \frac{1}{2}$ (Lemma 5.2)
   • If terminates at a Fano surface, use Cartier index bounds (Lemma 5.8)
4. Zariski decomposition (Proposition 4.7): Analyze types of contracted curves
5. Iterative improvement: Refine bounds through partial resolution and re-running MMP

Key lemmas:

• Lemma 5.8: Use Alexeev-Mori results to bound Cartier indices of Fano surfaces
• Lemmas 3.5-3.7: Adjunction formulas and partial resolution of log canonical centers

Effective Birationality

Theorem 5.28 (Explicit birationality constant): For $\epsilon < \tau_0$, $|m(K_{\mathcal{F}} + \epsilon K_X)|$ defines a birational map when:

• $K_X$ not pseudo-effective: $m \geq M(\eta,\epsilon) := 8!\left(2 + 8 \cdot \frac{1}{\epsilon} \cdot \left\lfloor 2\left(\frac{2\eta\epsilon}{\epsilon+1}\right)^{(2)^7/(\frac{\eta\epsilon}{\epsilon+1})^5}\right\rfloor!\right)$
• $K_X$ pseudo-effective with $\kappa(K_X) \neq 0$: $m > v(\eta) \cdot 42 \cdot 84^{128 \cdot 42^5 + 168}$

Proof method:

1. Volume lower bound method (Lemma 5.16): Using Birkar's results, if $\text{Vol}(K_{\mathcal{F}} + \epsilon K_X) \geq v$, then $m \geq \sqrt{v(\eta)/v}$ ensures birationality
2. Kodaira dimension classification:
   • $\kappa(K_X) = 2$: Use Alexeev-Mori to bound volume (Lemma 5.18)
   • $\kappa(K_X) = 1$: Apply Kawamata-Viehweg vanishing (Lemma 5.26)
   • $\kappa(K_X) = 0$: Surface is non-rational
   • $K_X$ not pseudo-effective: Terminates at Fano surface, use Kollár's results (Lemma 5.19)
3. Viehweg product technique (Lemma 5.23): For fibration cases, study fiber products $X^k := X \times_Y \cdots \times_Y X$

Experimental Setup

This is a pure theoretical mathematics paper (algebraic geometry) with no experimental setup, datasets, or computational experiments. All results are obtained through rigorous mathematical proofs.

Main Results

Summary of Theoretical Results

1. Universal lower bound for pseudo-effective thresholds: $\tau_0 = \frac{1}{3(2(2(3(2 \cdot 142151296)! + 1))^{128(3(2\cdot142151296)!+1)^5})!}$ This is an extremely small but explicit positive number
2. Birationality constants:
   • Fano case: $M_0(\epsilon) = 8!\left(2 + 8 \cdot \frac{1}{\epsilon} \cdot \left\lfloor 2\left(\frac{2\epsilon}{1+\epsilon}\right)^{(2)^7/(\frac{\epsilon}{1+\epsilon})^5}\right\rfloor!\right)$
   • General type case: $M_0 = 64 \cdot 193^2 \cdot 42 \cdot 84^{128 \cdot 42^5 + 168}$
3. Solution to Poincaré problem (Theorem 6.1): For foliations on $\mathbb{P}^2$ birationally equivalent to non-equivariant fibrations of genus $g \geq 2$, $\deg F \leq M_0(\tau_0)\left(\frac{1}{\tau_0} + 1\right)(4g-4) \cdot \frac{1}{\tau_0} \deg \mathcal{F}$
   Key improvement: This is a linear bound in $g$, whereas Pereira-Svaldi (2019) had exponential growth

Technical Results

1. Complete theory of adjoint MMP (Theorem 4.4):
   • Exists for any $\epsilon > 0$
   • Explicit two-phase decomposition
   • Singularity control: $\eta$-lc $\to$ $\frac{\epsilon\eta}{\epsilon+1}$-lc
2. DCC property (Theorem 5.5): Pseudo-effective threshold sets satisfy the descending chain condition
3. Cone and contraction theorems (Theorem 4.1): Extension of surface MMP theory to foliated case

Example Verification

Example 5.4: For a foliation $\mathcal{F}$ of degree $d$ on $\mathbb{P}^2$, $\tau = \frac{d-1}{3}$ since $K_{\mathcal{F}} = (d-1)H$ and $K_{\mathbb{P}^2} = -3H$, where $H$ is the hyperplane class. This verifies the correctness of the theory.

Related Work

Historical Development

1. Poincaré (1891): First posed the problem, proved that bounding the degree of general leaves suffices to determine existence of rational first integrals
2. Pereira (2002): Proved that the degree of general leaves of foliations of general type can be bounded, but provided no explicit bound
3. Pereira-Svaldi (2019):
   • Provided explicit bounds but with exponential dependence on genus
   • Introduced adjoint foliation framework
   • Posed Problem 1.1 as key to improving bounds
4. Spicer-Svaldi (2023):
   • Developed adjoint foliation theory for rank one foliations
   • Proved existence of adjoint MMP for $\epsilon \ll 1$
   • Proved DCC property of pseudo-effective thresholds (for $\epsilon \ll 1$)
5. Cascini-Spicer (2021, 2025):
   • Established MMP for three-dimensional foliations
   • Developed F-dlt correction theory
   • Adjunction formulas

Positioning of This Work

This paper is the first to provide an explicit answer to Problem 1.1, with main innovations:

1. Theory extension: Extends SS23 results from $\epsilon \ll 1$ to arbitrary $\epsilon > 0$
2. Explicit bounds: Provides explicit numerical expressions for all key constants
3. Linear improvement: Improves exponential bounds to linear bounds
4. Complete framework: Establishes complete theoretical chain from MMP to effective birationality

Related Techniques

1. MMP theory: Surface MMP by Kollár-Mori, Fujino
2. Fano surface theory: Bounds by Alexeev-Mori
3. Adjunction theory: Weak positivity by Viehweg
4. Effective birationality: ACC theory by Birkar, Hacon-McKernan-Xu

Conclusions and Discussion

Main Conclusions

1. Complete resolution of Poincaré problem: Provides explicit solution to the Poincaré problem in the Cerveau-Lins Neto formulation, with bounds linear in genus $g$
2. Complete theory of adjoint foliations:
   • Existence of adjoint MMP for arbitrary $\epsilon > 0$
   • DCC property of pseudo-effective thresholds
   • Explicit universal lower bound $\tau_0$
3. Effective birationality: For $\epsilon < \tau_0$, provides explicit constants $M_0(\epsilon)$ ensuring $|m(K_{\mathcal{F}} + \epsilon K_X)|$ is birational

Limitations

The author explicitly notes in the introduction:

1. Constants are far from optimal: "While the constants above are far from optimal"
   • $\tau_0 \approx 10^{-10^{10^9}}$ order of magnitude, extremely small
   • $M_0$ involves multiple factorials, extremely large
   • Better bounds likely exist in practice
2. Technical limitations:
   • Applies only to surfaces (dimension 2)
   • Requires $\epsilon$-adjoint log canonical singularity assumptions
   • Stronger results for $\Delta^{n-inv} = 0$ case
3. Method dependence:
   • Heavily relies on Alexeev-Mori bounds for Fano surfaces (themselves not optimal)
   • Uses Birkar's effective birationality results (involving large constants)
   • Cartier index bounds from combinatorial arguments (factorial growth)

Future Directions

Directions suggested by the paper:

1. Optimize constants:
   • Improve lower bounds for $\tau_0$
   • Reduce growth rate of $M_0$
   • Seek more refined analytical methods
2. Higher-dimensional generalizations:
   • Extend results to dimension three or higher
   • Combine with CHL+24 algebraically integrable foliation theory
   • Develop higher-dimensional adjoint MMP
3. Application extensions:
   • Study degree bounds for other foliation types
   • Applications to dynamical systems
   • Connections with moduli space theory
4. Theoretical deepening:
   • Understand precise nature of pseudo-effective thresholds
   • Study moduli spaces of $\epsilon$-adjoint singularities
   • Develop more refined classification theory

In-Depth Evaluation

Strengths

1. Theoretical Breakthrough

• Resolves long-standing open problem: Explicit version of Poincaré problem (1891)
• Key improvement: From exponential to linear bounds, a substantial advance
• Complete theory: Establishes complete framework from MMP to effective birationality

2. Technical Innovation

• Extended adjoint MMP: From $\epsilon \ll 1$ to arbitrary $\epsilon > 0$, significantly increased technical difficulty
• Singularity control: Proposition 3.10 provides precise control of base space singularities during MMP
• Decomposition technique: Lemma 3.8 MMP decomposition is key technical innovation
• Zariski decomposition analysis: Proposition 4.7 provides refined classification of contracted curves

3. Explicitness

• All constants explicit: Though not optimal, all are computable
• Verifiability: Theoretical predictions can be verified on concrete examples (e.g., Example 5.4)
• Algorithmic potential: In principle implementable as algorithms

4. Mathematical Rigor

• Complete proofs: All main results have detailed proofs
• Clear logic: Complete logical chain from basic definitions to final theorems
• Accurate references: Accurate citations and comparisons with prior work

Weaknesses

1. Practical Utility of Constants

• Extremely non-optimal: $\tau_0$ too small to use practically, $M_0$ too large to compute
• Missing improvement discussion: Lacks in-depth discussion of how to improve these constants
• Missing numerical examples: No computational examples for specific foliations

2. Technical Complexity

• High entry barrier: Requires deep background in algebraic geometry (MMP, foliation theory)
• Lengthy proofs: Some proof steps could be more concise
• Heavy notation: Extensive notation may impact readability

3. Result Limitations

• Dimension restriction: Applies only to surfaces
• Singularity assumptions: Requires technical $\epsilon$-adjoint lc assumptions
• Limited generality: Weaker results for $\Delta^{n-inv} \neq 0$ case

4. Application Demonstration

• Insufficient examples: Only one simple example ($\mathbb{P}^2$ foliations)
• Missing computations: No actual computation of degree bounds for specific foliations
• Lacking geometric intuition: Missing geometric interpretation of results

Impact Assessment

Contribution to Field

1. Milestone result: Resolves explicit version of Poincaré problem
2. Theoretical foundation: Establishes foundation for effective theory of foliations
3. Methodology: Systematic development of adjoint foliation framework

Practical Value

1. High theoretical value: Proves principle feasibility
2. Limited practical value: Constants too large for actual application
3. Inspirational value: Provides direction for improving bounds

Reproducibility

1. Theory reproducible: Complete proofs allow expert verification
2. Computation infeasible: Constants too large for practical computation
3. Concepts transferable: Methods applicable to other problems

Applicable Scenarios

Theoretical Research

• Foliation classification theory
• Birational geometry research
• Further MMP development

Potential Applications

• Dynamical systems: Understanding algebraic solutions of integrable systems
• Differential equations: Algebraic integrability determination
• Algebraic geometry: Moduli spaces of surfaces and foliations

Inapplicable Scenarios

• Practical computation: Constants too large
• Numerical simulation: Requires symbolic computation
• Low-precision estimates: Bounds too coarse

Technical Highlights in Detail

1. Two-Phase Decomposition of Adjoint MMP

Core of Theorem 4.4: $X \xrightarrow{f_0} X_0 \xrightarrow{f_1} X_1$

First phase $f_0$: Contracts $K_{\mathcal{F}} + \Delta^{n-inv}$-negative curves

• Geometric meaning: Handles foliation singularities
• Technical key: Uses Corollary 4.3 $K_{\mathcal{F}} + \Delta^{n-inv}$-MMP

Second phase $f_1$: Contracts curves that are $K_{X_0} + \Delta_0$-negative but $K_{\mathcal{F}_0} + \Delta_0^{n-inv}$-positive

• Geometric meaning: Adjusts base space geometry
• Technical key: Uses Fujino's surface MMP

Innovation: This decomposition allows independent control of foliation and base space singularities, key to extending to arbitrary $\epsilon > 0$.

2. DCC Proof for Pseudo-effective Thresholds

Proof strategy of Theorem 5.5:

1. Run $K_{(X,\mathcal{F},\Delta)_\tau}$-MMP to obtain semi-ample divisor
2. Apply adjunction formula on general fiber
3. Use fiber boundedness to obtain ACC property for $\frac{1}{\tau}$

Key observation: $\frac{1}{\tau} = \frac{1}{d + \sum a_j b_j} \cdot \left(c - \sum a_j b_j\right)$ where $d, b_j, c$ are bounded integers and coefficients $a_j \in \Delta$ belong to DCC set.

Technical elegance: Converts geometric problem (pseudo-effectiveness) to combinatorial problem (rational combinations of bounded integers).

3. Cartier Index Control for Fano Surfaces

Application of Lemma 5.8:

• Use Alexeev-Mori: Picard number $\rho \leq \frac{128}{\eta^5}$ for Fano surfaces
• Intersection matrix determinant $\det(E_{ij}) \leq 2 \cdot \left(\frac{2}{\eta}\right)^{128/\eta^5}$
• Cartier index divides $N = \left\lfloor 2 \cdot \left(\frac{2}{\eta}\right)^{128/\eta^5}\right\rfloor!$

Geometric intuition: $\eta$-lc singularities limit complexity of Fano surfaces, hence limit Cartier indices.

Constant origin: Multiple factorial appearance is unavoidable due to combinatorial arguments.

Key References

1. Per02 Pereira, J. V. (2002). Poincaré problem for foliations of general type. Math. Ann., 323(2):217-226.
   • First proved degree boundedness, but non-explicitly
2. PS19 Pereira, J. V., Svaldi, R. (2019). Effective algebraic integration in bounded genus. Algebr. Geom., 6(4):454-485.
   • Posed Problem 1.1, provided exponential bounds
3. SS23 Spicer, C., Svaldi, R. (2023). Effective generation for foliated surfaces: results and applications. J. Reine Angew. Math., 795:45-84.
   • Established $\epsilon \ll 1$ adjoint foliation theory
4. CS21 Cascini, P., Spicer, C. (2021). MMP for co-rank one foliations on threefolds. Invent. Math., 225(2):603-690.
   • F-dlt correction theory
5. AM04 Alexeev, V., Mori, S. (2004). Bounding singular surfaces of general type.
   • Fano surface bounds, main source of constants in this paper

Summary

This paper represents an important theoretical breakthrough in birational geometry of foliations, completely resolving the explicit version of the Poincaré problem and improving bounds from exponential to linear. While specific constants are non-optimal, it proves principle feasibility and establishes solid foundation for further development in the field. The paper achieves high levels of technical depth and theoretical completeness, representing excellent work in algebraic geometry.

Recommended for:

• Algebraic geometers (especially MMP and foliation specialists)
• Scholars interested in the Poincaré problem
• Mathematicians researching effective algebraic geometry

Reading suggestions:

• Requires solid algebraic geometry background (MMP, singularity theory)
• Recommend first reading PS19 and SS23 for background
• Focus on proof strategies rather than specific constant values