Abstract:Let $S$ be a very general smooth hypersurface of degree $6$ in $\mathbb{P}^3$. In this paper we will prove that the moduli space of $\mu$-stable rank $2$ torsion free sheaves with respect to hyperplane section having $c_1 = \mathcal{O}_S(1)$, with fixed $c_2 \ge 27$ is irreducible.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is whether the rank - 2 μ - stable torsion - free sheaf moduli space is irreducible under specific conditions on a very general sextic surface. Specifically, the author studies the irreducibility of the moduli space \(M(H, c_2)\) when the second Chern class \(c_2\geq27\). The main conclusion of the paper is that for \(c_2\geq27\), the moduli space \(M(H, c_2)\) is irreducible.
### Background and Motivation
1. **Construction of the Moduli Space**
- The moduli space \(M(H, c_2)\) was constructed by Gieseker and Maruyama using Mumford's geometric invariant theory. It is a projective scheme (not necessarily reduced).
- Many researchers have studied the geometric properties of the moduli spaces over different bases, such as rational surfaces, K3 surfaces, etc.
2. **Previous Work**
- When the base is a quintic hypersurface, Simpson and Mestrano systematically studied this problem and proved that the rank - 2 torsion - free sheaf moduli space is generically smooth and irreducible under certain conditions.
- Nijsse proved in his unpublished work that for \(c_2\geq16\), the moduli space is irreducible.
3. **Raising of the Problem**
- Simpson and Mestrano proved that on a very general sextic surface, when \(c_2 = 11\), the moduli space is not irreducible.
- The natural question is whether there is an effective lower bound for \(c_2\) such that the moduli space becomes irreducible.
### Main Results
- **Theorem 1.1**: For \(c_2\geq27\), the moduli space \(M(H, c_2)\) is irreducible.
### Proof Ideas
1. **Deformation to the Boundary**
- The author uses O'Grady's method to prove the irreducibility of the moduli space by deforming to the boundary.
- The key to the proof is the Connecting Theorem, that is, the connectedness of \(M(H, c_2)\).
2. **Non - emptiness**
- The author proves that certain Harder - Narasimhan sheaves (HN sheaves) are non - empty, which is a crucial step in proving the connectedness.
3. **Existence of the Boundary**
- The author proves that when \(c_2\geq34\), any irreducible good component \(X\subset M(H, c_2)\) has a non - empty boundary.
### Technical Details
- **Kuranishi Theory**
- Through Kuranishi theory, the author proves that \(M(H, c_2)\) is a local complete intersection when \(c_2\geq27\).
- **Hartshorne's Connectivity Theorem**
- Hartshorne's Connectivity Theorem states that if two different irreducible components intersect at a point, then they intersect in a sub - variety of codimension 1.
- **Analysis of Singular Points**
- The author analyzes in detail the singular points of the moduli space, especially those vector bundles with non - zero sections.
### Conclusion
The paper finally proves, through a series of technical means, that when \(c_2\geq27\), the moduli space \(M(H, c_2)\) is irreducible. This result provides an important theoretical basis for understanding the moduli spaces of vector bundles on very general sextic surfaces.