Interpolation and moduli spaces of vector bundles on very general blowups of the projective plane

Basic Information

• Paper ID: 2306.06175
• Title: Interpolation and moduli spaces of vector bundles on very general blowups of the projective plane
• Authors: Izzet Coskun (University of Illinois at Chicago), Jack Huizenga (Pennsylvania State University)
• Classification: math.AG (Algebraic Geometry)
• Publication Date: 2024, Épijournal de Géométrie Algébrique, Volume 8, Article No. 7
• Paper Link: https://arxiv.org/abs/2306.06175

Abstract

This paper investigates certain moduli spaces of vector bundles on blowups of the projective plane P² at at least 10 very general points. Sheaf moduli spaces on surfaces of general type can be non-reduced, reducible, or even disconnected. In contrast, moduli spaces on minimal rational surfaces and certain del Pezzo surfaces are irreducible and smooth on the stable bundle locus. The authors discover examples of moduli spaces of vector bundles on more general blowups of P² that are disconnected and possess branches of different dimensions. In fact, assuming the SHGH conjecture holds, one can find moduli spaces with arbitrarily many branches of arbitrarily large dimension.

Research Background and Motivation

1. Core Problems to Address:
   • Study the geometric structure of moduli spaces of vector bundles on surfaces X obtained by blowing up P² at multiple general points
   • Focus on the connectivity and branch structure of MX,At(2,K,χ) when the number of blowup points n≥10
2. Importance of the Problem:
   • Topological properties of moduli spaces are crucial for understanding moduli theory in algebraic geometry
   • The behavior of sheaf moduli spaces on rational surfaces exhibits significant differences under various conditions
   • These investigations provide guidance for understanding moduli spaces on more general surfaces
3. Limitations of Existing Research:
   • For minimal rational surfaces and certain del Pezzo surfaces, moduli spaces are known to be irreducible and smooth on the stable bundle locus
   • Walter proved that moduli spaces are irreducible under certain conditions when (KY + F)·A < 0
   • However, the behavior of moduli spaces on more general rational surfaces, particularly when Walter's condition is violated, remains unclear
4. Research Motivation:
   • Explore pathological behaviors that may occur in moduli spaces on rational surfaces
   • Determine under what conditions reducible and disconnected moduli spaces appear
   • Provide new examples and counterexamples for a general theory of topological properties of moduli spaces

Core Contributions

1. Discovery of Pathological Phenomena Observed for the First Time on Rational Surfaces:
   • Constructed disconnected moduli spaces of vector bundles with branches of different dimensions
   • Under the assumption of the SHGH conjecture, proved that moduli spaces with arbitrarily many branches of arbitrarily large dimension can be constructed
2. Established Classification Theory for Bundles:
   • Proved that every vector bundle V with positive Euler characteristic has a unique "type D"
   • Provided complete classification of type D, particularly for 10≤n≤17
3. Provided Concrete Descriptions of Moduli Spaces:
   • For n=16: MX,At(2,K,2) ≅ P⁵ (when 14/3 < t < 16/3); blowup of P⁵ at 16 points (when 4 < t < 14/3)
   • For n=25: MX,At(2,K,4) ≅ disjoint union of 25 copies of P⁸ (when 5 < t ≤ 27/5)
4. Established Connection with Continued Fraction Expansions:
   • When n is not a perfect square, classification of effective divisors D can be achieved by solving generalized Pell equations
   • Transformed geometric problems into number-theoretic problems, providing computational tools

Detailed Methodology

Problem Setup

Study moduli spaces MX,At(r,c₁,χ) of vector bundles on the surface X = Bl_{p₁,...,pₙ}P², where:

• X is the blowup of P² at n very general points
• At = tH - E is the polarizing divisor (t > √n)
• Focus on the case (r,c₁,χ) = (2,K,χ), where K is the canonical divisor

Core Theoretical Framework

1. Bundle Type Classification

Definition 3.1: Let v = (2,K,χ) with χ≥1, and D an effective divisor class satisfying χ(D)≥1. A vector bundle V has type D if it fits into an exact sequence:

0 → O(D) → V → K(-D)⊗I_Z → 0

where Z is a zero-dimensional scheme of length 2χ(O(D))-χ.

Theorem 3.2: Every vector bundle V with Chern character v = (2,K,χ) (χ≥1) has exactly one type D.

2. Stability Conditions

Proposition 3.9: If V is a bundle of type D and there exists a polarization At₀ such that V is μ-At₀ semistable, then:

• 2B·D < B·K
• There exists a unique polarization AtD such that 2AtD·D = AtD·K

3. Classification of Effective Divisors

For effective divisors D satisfying 2B·D < B·K and χ(D)≥1:

When 10≤n≤12 (Theorem 4.8): Divisors D are related to odd convergents of the continued fraction expansion of √n:

• For positive odd k, define dk = ½(pk-3), mk = ½(qk-1)
• Dk = dkH - mkE is a divisor satisfying the conditions

When n=16 (Theorem 4.13): The only effective divisors satisfying the conditions are O and the exceptional divisors Ei.

Technical Innovations

1. Establishment of Type Theory:
   • Established unique decomposition of bundles through analysis of exact sequences
   • Transformed geometric problems into cohomological computation problems
2. Application of Number-Theoretic Methods:
   • Utilized generalized Pell equations to solve divisor classification problems
   • Established deep connections between geometric objects and continued fractions
3. Precise Description of Moduli Spaces:
   • Provided exact dimensions of various branches through analysis of extension classes
   • Proved that branches corresponding to different types are disjoint

Experimental Setup

Main Computational Tools

1. Application of SHGH Conjecture: Assume the Segre-Harbourne-Gimigliano-Hirschowitz conjecture holds
2. Nagata Conjecture: When n≥10 and t>√n, At is ample
3. Cohomological Computations: Utilize Riemann-Roch theorem and exact sequences of sheaves

Concrete Case Analysis

Case 1: n=10

Effective divisors D form an infinite sequence:

O, 57H-18E, 2220H-702E, 84357H-26676E, ...

These divisors can be read from the continued fraction expansion of √10.

Case 2: n=13

Divisors fall into 6 infinite families, each corresponding to different balancing conditions:

• Family (I): O, 2142H-594E, ...
• Family (II): 195H-54E, 255057H-70740E, ...
• And so on

Experimental Results

Main Results

Theorem 1.1 (Concrete Moduli Space Descriptions)

1. n=16:
   • When 14/3 < t < 16/3: MX,At(2,K,2) ≅ P⁵
   • When 4 < t < 14/3: MX,At(2,K,2) ≅ blowup of P⁵ at 16 points
2. n=25:
   • When 5 < t ≤ 27/5: MX,At(2,K,4) ≅ disjoint union of 25 copies of P⁸

Theorem 1.3 (General Structure, Assuming SHGH Conjecture)

For 10≤n≤15:

1. When t > n/3, MX,At(2,K,2) is empty
2. As t decreases through n/3, a branch isomorphic to P^(n-11) is acquired
3. For each non-trivial divisor D, as t decreases through tD, a new branch of dimension -χ(2D-K)-1 is acquired
4. All branches are mutually disjoint

Topological Complexity Results

Corollary 1.4: Assuming the SHGH conjecture, for 10≤n≤12, given positive integers k and r, there exists ε>0 such that when √n < t < √n+ε, the moduli space MX,At(2,K,χ) has at least k irreducible components of dimension r.

Numerical Examples

For the case n=12, the wall-crossing sequence is:

D           tD      New Branch
O           4       P¹
21H-6E      52/15   P¹⁰
312H-90E    724/209 P¹⁴⁵
...

Related Work

Classical Results

1. Walter's Theorem: When (KY + F)·A < 0, moduli spaces on birationally ruled surfaces are irreducible
2. O'Grady's Theorem: As χ tends to negative infinity, moduli spaces become irreducible, reduced, and normal
3. Göttsche's Results: Betti numbers of Hilbert schemes increase monotonically with n

Contributions of This Paper in Comparison

• First discovery of disconnected moduli spaces on rational surfaces
• Violates the general expectation of monotonicity of Betti numbers
• Provides concrete examples when Walter's condition fails

Conclusions and Discussion

Main Conclusions

1. Moduli spaces on rational surfaces can exhibit complex topological behavior
2. When Walter's condition is violated, irreducibility may fail
3. The validity of the SHGH conjecture is crucial for understanding these phenomena

Limitations

1. Main results depend on the SHGH conjecture (except in perfect square cases)
2. Analysis is primarily restricted to specific Chern characters
3. Computational complexity grows rapidly with increasing n

Future Directions

1. Extension to higher rank vector bundles
2. Investigation of similar phenomena on other types of surfaces
3. Development of methods not depending on the SHGH conjecture

In-Depth Evaluation

Strengths

1. Theoretical Innovation: First discovery of pathological moduli spaces on rational surfaces, challenging existing knowledge
2. Methodological Completeness: Combines multiple tools from algebraic geometry, number theory, and cohomological theory
3. Computational Precision: Provides exact dimensions and precise descriptions of moduli spaces
4. Richness of Examples: Offers multiple concrete numerical examples and classifications

Weaknesses

1. Conjecture Dependence: Core results depend on the SHGH conjecture, limiting unconditional validity
2. Limited Scope: Focus on specific Chern characters with limited generalization
3. Computational Complexity: Classification becomes extremely complex for large values of n

Impact

1. Theoretical Contribution: Provides important counterexamples and new phenomena for moduli space theory
2. Methodological Value: The established technical framework can be applied to other geometric problems
3. Practical Significance: Provides guidance for understanding moduli spaces on more general surfaces

Applicable Scenarios

This methodology is suitable for application to:

1. Study of moduli spaces on rational surfaces
2. Problems requiring precise computation of moduli space branches
3. Applications of continued fraction methods in algebraic geometry

References

The paper cites 29 important references, primarily including:

• Classical work by Nagata on ampleness
• Literature on the SHGH conjecture
• Foundational literature on moduli space theory (Gieseker, Maruyama, etc.)
• Related work on topological properties (O'Grady, Göttsche, etc.)