Abstract: In previous work by Coates, Galkin and the authors, the notion of mutation between lattice polytopes was introduced. Such mutations give rise to a deformation between the corresponding toric varieties. In this paper we study one-step mutations that correspond to deformations between weighted projective planes, giving a complete characterization of such mutations in terms of T-singularities. We also show that the weights involved satisfy Diophantine equations, generalizing results of Hacking and Prokhorov.
Introduction
In algebraic geometry, toric varieties are often studied through the lens of convex lattice polytopes. A "mutation" is a combinatorial operation on these polytopes that corresponds to a geometric "deformation" of the associated variety. While the general theory of mutations was established in previous work, this paper focuses on a specific class: weighted projective planes (WPPs) and their quotients, known as fake weighted projective planes.
The central challenge is to determine exactly when one WPP can be mutated into another. This involves understanding the relationship between the "weights" (the integers defining the projective space) and the singularities present on the variety. The authors aim to bridge the gap between the combinatorial rules of mutation and the number-theoretic properties of the varieties' weights.
Methodology
The authors employ a mix of combinatorial geometry, singularity theory, and number theory. WPPs are represented as triangles in a 2D lattice. Mutation is defined as a piecewise linear transformation of these triangles, often visualized through the movement of vertices within the lattice grid.
The study focuses on $T$-singularities—quotient surface singularities that admit a particular kind of smoothing (Q-Gorenstein). Furthermore, the authors map the weights of the projective planes to solutions of generalized Markov equations of the form: $$mx_0x_1x_2 = k(c_0x_0^2 + c_1x_1^2 + c_2x_2^2)$$ Properties of these equations, such as Pell equations, are used to explore the structure of "mutation trees."
Key Insights: The T-Singularity Criterion
The most significant insight is that a one-step mutation exists if and only if the edge of the triangle being mutated corresponds to a $T$-singularity. This provides a rigid geometric condition for a purely combinatorial process.
If a WPP has weights $(\lambda_0, \lambda_1, \lambda_2)$, a mutation across the edge opposite $\lambda_0$ results in a new weight: $$\lambda_{new} = \frac{(\lambda_1 + \lambda_2)^2}{\lambda_0}$$ This formula ensures that the "degree" of the variety remains invariant under mutation. The authors show that mutation is equivalent to a standard transformation used to find new solutions to Markov-type Diophantine equations.
Key Results
The paper establishes several foundational results for the classification of these varieties:
- Geometric Characterization: Proposition 1.1 provides the definitive characterization: a one-step mutation from $X$ to $Y$ exists if and only if the weights satisfy the $T$-singularity condition.
- Invariance of Multiplicity: Mutation preserves the "multiplicity" of a fake weighted projective plane. If one starts with a quotient of a WPP, the resulting variety is a quotient of the same order.
- Classification of Solutions: The paper demonstrates how to generate all possible weighted projective planes that are mutation-equivalent by starting from "minimal" weights and applying mutations, effectively mapping out a "forest" of mutation trees.
Limitations and Future Work
While this work provides a complete characterization for 2D surfaces (triangles), extending these results to higher-dimensional Fano varieties remains a complex open problem. The current study focuses on "one-step" mutations; while these are the building blocks, multi-step mutations or more general mutations not involving triangles are not yet fully characterized. Finally, while these mutations are deeply related to Mirror Symmetry (specifically mutations of Laurent polynomials), the explicit mapping to the "A-side" of the mirror remains a broader research goal.
Summary updated