Mirzakhani–Wright Rank Obstruction

• The Mirzakhani–Wright rank obstruction is a geometric constraint on GL(2,R)-invariant orbit closures, quantifying deformation limits via rank and homological dimension.
• It shows that degeneration through pinching absolute cycles enforces a drop in rank, thus restricting the types of deformations in moduli spaces.
• Applications include the classification of Veech surfaces in rational polygons and determining the number of zero Lyapunov exponents in the Kontsevich–Zorich spectrum.

The Mirzakhani–Wright rank obstruction is a fundamental phenomenon in the geometry and dynamics of $GL(2,\mathbb{R})$–invariant subvarieties of strata of Abelian differentials. It stipulates that the ability of an orbit closure to deform in absolute cohomology is intrinsically limited by the topological complexity of those deformations, quantified via the rank and homological dimension. This obstruction plays a pivotal role in classifying invariant subvarieties, constraining degenerations, and determining the number of zero Lyapunov exponents in the Kontsevich–Zorich spectrum. Its arithmetic reformulation yields powerful applications in the classification of Veech surfaces, particularly in the context of rational polygons, such as rational triangles (Apisa, 2021, Chen et al., 2019, Angdinata et al., 25 Mar 2026).

1. Definitions: Rank, Rel, and Homological Dimension

Let $\Omega\mathcal{M}_g$ be the bundle of holomorphic one-forms over the moduli space $\mathcal{M}_g$ of genus-$g$ curves. A $GL(2,\mathbb{R})$–invariant orbit-closure $\mathcal{M} \subset \Omega\mathcal{M}_g$ (an affine invariant subvariety) has tangent spaces $T_{(X,\omega)}\mathcal{M}$ canonically identified with complex linear subspaces of the relative cohomology $H^1(X,\Sigma;\mathbb{C})$, where $\Sigma$ is the zero locus of $\omega$. Projection to absolute cohomology,

$\Omega\mathcal{M}_g$0

our focus is on the image $\Omega\mathcal{M}_g$1, a symplectic subspace. The rank is defined by

$\Omega\mathcal{M}_g$2

while the dimension of the kernel, denoted $\Omega\mathcal{M}_g$3, satisfies

$\Omega\mathcal{M}_g$4

The homological dimension, or “cylinder dimension,” is defined for horizontally periodic surfaces $\Omega\mathcal{M}_g$5 in $\Omega\mathcal{M}_g$6 as the maximal real dimension of the space spanned by core curves of horizontal cylinders in absolute homology,

$\Omega\mathcal{M}_g$7

Inequalities $\Omega\mathcal{M}_g$8 hold, linking the geometric and cohomological deformation theory (Apisa, 2021, Chen et al., 2019).

2. The Rank Obstruction: Tangent Space Formulation

Mirzakhani and Wright established a precise relationship between degenerations of orbit closures and the drop in rank at the boundary. For a sequence $\Omega\mathcal{M}_g$9 in a stratum $\mathcal{M}_g$0 converging to a boundary surface $\mathcal{M}_g$1 in the partial compactification $\mathcal{M}_g$2, the tangent space to the boundary stratum is the intersection

$\mathcal{M}_g$3

where $\mathcal{M}_g$4 is the subspace of vanishing cycles and $\mathcal{M}_g$5 is its annihilator in relative cohomology. The image under projection satisfies

$\mathcal{M}_g$6

with $\mathcal{M}_g$7. Hence,

$\mathcal{M}_g$8

This is the Mirzakhani–Wright rank obstruction: if degeneration kills a nontrivial absolute cycle, rank must drop, so maximal-rank orbit closures cannot degenerate by pinching such cycles (Chen et al., 2019).

3. Classification of Minimal Homological Dimension and Cylinder-Homology Criterion

A subvariety $\mathcal{M}_g$9 has minimal homological dimension when $g$0. Mirzakhani and Wright proved that under this assumption, the only possibilities are:

• $g$1 is a connected component of a stratum of Abelian differentials,
• or $g$2 is the full locus of degree-$g$3 covers of the hyperelliptic locus in some stratum.

Moreover, the minimal dimension condition is equivalent to a “cylinder-homology” property: for every $g$4 in $g$5, any two $g$6-parallel cylinders (those in the same equivalence class under cylinder deformations) have homologous core curves. In cohomological terms, for every equivalence class, the “standard shear” supported on that class projects nontrivially to a one-dimensional subspace, enforcing strong homological rigidity (Apisa, 2021).

4. Rank Obstruction and Lyapunov Exponents

Forni’s criterion links the number of nonzero Lyapunov exponents for the Kontsevich–Zorich cocycle to $g$7, with the extremal upper bound on the number of zero Lyapunov exponents,

$g$8

Equality is achieved if and only if $g$9. The only non-full-rank affine invariant subvarieties meeting this bound are the Teichmüller curves associated to the Eierlegende-Wollmilchsau (genus 3) and Ornithorynque (genus 4). Outside these exceptions and full-rank loci, the obstruction precludes invariant subvarieties from achieving the maximal count of zero exponents (Apisa, 2021).

5. Applications to Translation Surfaces and Polygonal Billiards

The rank obstruction provides a geometric-analytic classification criterion for translation surfaces and their orbit closures. In particular, for Veech (lattice) surfaces arising from rational polygons, the Mirzakhani–Wright rank obstruction rules out the existence of lattice triangles in certain regimes. Specifically, the Teichmüller curve condition ($GL(2,\mathbb{R})$0) requires that the rank be exactly one. For rational triangles, a number-theoretic reformulation shows that for almost all triangles in the “hard obtuse window” (largest angle in $GL(2,\mathbb{R})$1), the rank obstruction is satisfied, meaning they are not lattice surfaces. The density of such exceptions tends to zero as the denominator grows, provided it has a suitably large prime factor (Angdinata et al., 25 Mar 2026).

Surface type	Rank required for Teichmüller curve	Rank obstruction consequence
Veech surface/Teichmüller curve	1	$GL(2,\mathbb{R})$2 must have $GL(2,\mathbb{R})$3
Full-rank stratum	$GL(2,\mathbb{R})$4	Cannot degenerate by pinching absolute cycles

6. Degeneration, Compactification, and Monodromy Arguments

Partial compactification $GL(2,\mathbb{R})$5 plays a critical role: degeneration by pinching cylinders or vanishing cycles constrains the allowable orbit closures by the rank drop determined through the tangent space intersection with vanishing cycle annihilators. The existence of an “optimal translation cover” compatible with these degenerations, and monodromy arguments (Avila–Eskin–Möller; Filip), ensure that the complement of the image of $GL(2,\mathbb{R})$6 in absolute cohomology is the only possible source of zero Lyapunov exponents. Boundary strata reveal that true full-rank affine invariant subvarieties are severely restricted in their degenerations, as they cannot pinch nontrivial absolute cycles (Apisa, 2021, Chen et al., 2019).

7. Arithmetic Reformulation and Machine-Formalized Analysis

For rational triangles, the rank obstruction is turned into explicit modular inequalities: if there exists a “usable” $GL(2,\mathbb{R})$7 modulo the denominator $GL(2,\mathbb{R})$8 satisfying specific inequalities on the residues of the angles, then the associated translation surface fails to be a lattice surface. This arithmetic criterion admits Fourier-theoretic and Ramanujan-sum analysis, yielding a density-zero result for hard-obtuse rational triangles. The proof and its core analytic arguments—particularly the main-term/error-term dichotomy and large-prime suppression of error—have been fully autoformalized in Lean 4 using AxiomProver, validating the arguments behind the density theorem rigorously (Angdinata et al., 25 Mar 2026).

8. Examples and Corollaries

• Classical lattice triangles exist in the acute and right-angled regimes, classified by Kenyon–Smillie and Schlage–Puchta.
• Obtuse regime contains only two infinite families and a single sporadic example.
• Hard-obtuse window: No new examples known; Mirzakhani–Wright’s obstruction, along with the modular arithmetic criterion, explains the observed paucity.
• Geometric implications: Full-rank subvarieties cannot contain arbitrarily many cylinders of unbounded modulus, enforcing finiteness in surface decompositions and moduli dynamics (Apisa, 2021, Angdinata et al., 25 Mar 2026).

In summary, the Mirzakhani–Wright rank obstruction unifies geometric, cohomological, and arithmetic considerations, imposing sharp constraints on the deformation theory of translation surfaces, the structure of orbit closures, and the arithmetic of polygonal billiards. Its ramifications are central to the modern stratification and ergodic theory of flat surfaces and their moduli (Apisa, 2021, Chen et al., 2019, Angdinata et al., 25 Mar 2026).