Tautological relations and integrable systems

Imagine you are trying to understand the shape of a vast, infinite landscape made of mathematical objects called curves. Specifically, these are "stable algebraic curves"—think of them as flexible, rubbery loops that can have holes (genus) and little stickers (marked points) attached to them. Mathematicians call the collection of all possible shapes these curves can take a Moduli Space.

This paper is like a detective story where two mathematicians, Alexandr and Sergey, are trying to find hidden rules (called relations) that govern how these shapes interact. They believe these rules are the "secret code" that connects two very different worlds: Geometry (the shapes of curves) and Physics (specifically, systems that evolve over time, known as Integrable Systems).

Here is the breakdown of their discovery using simple analogies:

1. The Two Worlds: Geometry vs. Physics

The Geometry World (The Curves): Imagine a giant Lego set where every piece is a different shape of a curve. Mathematicians have built a "Tautological Ring," which is like a rulebook describing how you can combine these Lego pieces.
The Physics World (The Integrable Systems): Imagine a complex machine with many gears (equations) that move in a perfectly synchronized way. If you turn one gear, the whole machine moves predictably. There are two famous machines in this paper:
- The DZ Machine (Dubrovin–Zhang): A very powerful machine, but its instructions are written in a messy, complicated language. Sometimes, the instructions look like they have "fractional" or broken parts that shouldn't exist.
- The DR Machine (Double Ramification): A newer machine built with cleaner, simpler instructions. It's easier to read, but it's hard to prove it does the exact same job as the DZ machine.

2. The Big Mystery: Are They the Same?

For decades, mathematicians suspected that the DZ and DR machines were actually the same machine, just wearing different masks. They believed there was a "translation manual" (called a Miura transformation) that could turn the messy DZ instructions into the clean DR instructions.

However, no one could prove that the messy DZ instructions were actually "clean" (polynomial) or that the translation manual worked for every possible scenario.

3. The New Discovery: The "Tree" Rules

The authors propose a new family of rules (conjectures) written in the language of the Geometry world.

The Analogy: Imagine you are trying to balance a stack of plates. The authors found that if you arrange the plates in a specific way—specifically, in the shape of trees (branches with no loops) and decorate them with simple numbers—they should perfectly balance out to zero.
The "Tree" Structure: They say, "If you take all these tree-shaped arrangements of curves and add them up with alternating signs (plus, minus, plus...), the total weight should be zero."
Why Trees? Trees are simple. They don't have loops. The authors found that the most complex, messy parts of the geometry can be broken down into these simple tree structures.

4. What Happens When the Rules Are True?

The paper argues that if these "Tree Rules" are true, then three huge things happen in the Physics world:

The Messy Machine Gets Clean: The DZ machine's instructions are proven to be perfectly clean (polynomial). The "fractional" parts vanish.
The Translation Works: The DZ and DR machines are proven to be the same. You can translate one to the other without losing any information.
The Secret Code is Cracked: They provide a specific geometric formula for how to translate between the two machines. It's like finding the exact key that unlocks the door between the two rooms.

5. The Proof: Solving the Puzzle

The authors didn't just guess; they proved their rules work in two specific, difficult scenarios:

Case 1: Only One Sticker (n=1): They proved the rules work when the curve has just one marked point, regardless of how many holes it has. They used a technique called Localization, which is like shining a special light on the curve landscape. When you shine this light, most of the landscape disappears, and you only see the "fixed points" (the most stable spots). By analyzing these spots, they could prove the rules hold.
Case 2: No Holes (g=0): They proved the rules work for curves with no holes (like simple circles or spheres), regardless of how many stickers are on them.

6. The "Aha!" Moment

The most exciting part is that by proving these rules for the simple cases, they confirmed a major conjecture from their previous work. They showed that the "Tree Rules" are the fundamental building blocks. If you understand how these simple trees balance, you understand the entire complex system of curves and physics.

Summary

Think of this paper as finding the Universal Grammar for a language spoken by both shapes and time-evolving systems.

The Problem: We have two dialects (DZ and DR) that we think are the same, but we can't prove it because one dialect is messy.
The Solution: The authors found a set of "Tree Sentences" (conjectural relations) that are so fundamental that if they are true, they force the messy dialect to become clean and prove it matches the other dialect.
The Result: They proved these sentences are true for the simplest versions of the language, giving us high confidence that the whole system works as they predict.

In short, they found a simple, tree-like pattern in the complex world of curves that explains why two major mathematical machines are actually the same thing.

Here is a detailed technical summary of the paper "Tautological relations and integrable systems" by Alexandr Buryak and Sergey Shadrin.

1. Problem Statement and Context

The paper addresses the deep connection between the geometry of the moduli spaces of stable algebraic curves, $\mathcal{M}_{g,n}$ , and integrable systems. Specifically, it focuses on two major hierarchies associated with Cohomological Field Theories (CohFTs) and their generalization, F-Cohomological Field Theories (F-CohFTs):

The Dubrovin–Zhang (DZ) Hierarchy: Constructed via the topological solution of the CohFT. While known to exist, a fundamental open problem has been the polynomiality of its equations (i.e., whether the right-hand sides are differential polynomials) and the lack of an explicit geometric formula for these equations.
The Double Ramification (DR) Hierarchy: Constructed using DR cycles. By construction, its equations are polynomial, but identifying a specific solution with a simple geometric interpretation (analogous to the DZ topological solution) has been challenging.

A central conjecture in the field posits that the DZ and DR hierarchies associated with the same theory are Miura equivalent (related by a specific change of variables). Previous work established this for standard CohFTs under certain conditions, but the general case for F-CohFTs and the precise nature of the equivalence remained open.

The authors propose a new family of conjectural tautological relations in the cohomology of $\mathcal{M}_{g,n+m}$ to resolve these issues.

2. Methodology

The authors employ a combination of combinatorial geometry, localization techniques, and the theory of integrable systems:

Combinatorial Formulation (Stable Trees): The core of the conjecture involves defining classes $B^m_{g,d}$ using sums over stable rooted trees. These trees are decorated with powers of $\psi$ -classes (Chern classes of cotangent line bundles) at half-edges. The relations are formulated as sums over these trees vanishing (or equating to specific DR-cycle expressions) depending on the parameter $m$ .
Localization in Moduli Spaces: To prove the conjectures in specific cases ( $n=1$ $n = 1$ and $g=0$ $g = 0$ ), the authors utilize the Liu–Pandharipande method. This involves:
- Considering the moduli space of stable relative maps to $(\mathbb{P}^1, \infty)$ .
- Applying equivariant localization with respect to a $\mathbb{C}^*$ -action on $\mathbb{P}^1$ .
- Analyzing the fixed-point loci, which correspond to decorated bipartite graphs, and computing contributions from virtual normal bundles.
- Extracting relations in the cohomology of $\mathcal{M}_{g,n}$ by equating coefficients of the equivariant parameter.
Reduction Arguments: The paper demonstrates that the infinite system of conjectural relations can be reduced to a finite set of relations of a specific degree, making the verification problem tractable.

3. Key Contributions

A. The Conjectural Relations

The authors introduce three families of conjectures parameterized by an integer $m \geq 0$ :

Conjecture 1 ( $m \geq 2$ ): The class $B^m_{g,d}$ $B_{g, d}^{m}$ vanishes for $\sum d_i \geq 2g + m - 1$ $\sum d_{i} \geq 2 g + m - 1$ .
- Significance: This implies the polynomiality of the DZ hierarchy equations for arbitrary F-CohFTs.
Conjecture 2 ( $m = 1$ ): The class $B^1_{g,d}$ $B_{g, d}^{1}$ equals a specific class $A^1_{g,d}$ $A_{g, d}^{1}$ constructed from DR cycles and Hodge classes.
- Significance: This implies that the DZ and DR hierarchies are related by a Miura transformation.
Conjecture 3 ( $m = 0$ ): The class $B^0_{g,d}$ $B_{g, d}^{0}$ equals a class $A^0_{g,d}$ $A_{g, d}^{0}$ (previously proposed in [BGR19]).
- Significance: This implies the normal Miura equivalence of the DZ and DR hierarchies, preserving the $\tau$ -structure.

B. Proofs of the Conjectures

The authors prove these conjectures in two fundamental cases:

Theorem 2.2: The conjectures hold for $n=1$ (arbitrary genus $g$ ). This is a significant result as it proves the main conjecture from their previous joint work with Hernández Iglesias. The proof relies on a new tautological relation derived via localization.
Theorem 2.3: The conjectures hold for $g=0$ (arbitrary number of marked points $n$ ). The proof utilizes the fact that the cohomology of $\mathcal{M}_{0,n}$ is generated by boundary strata, allowing a reduction to specific CohFT constructions.

C. Geometric Formulas for the DZ Hierarchy

A major theoretical breakthrough is the derivation of an explicit geometric formula for the polynomial part of the DZ hierarchy equations.

The authors show that the coefficients of the DZ hierarchy equations can be expressed as intersection numbers on $\mathcal{M}_{g,n}$ involving the universal cohomology classes $B^m_{g,d}$ and the F-CohFT classes.
Since the polynomiality is proven for semisimple CohFTs, this provides a concrete geometric description of the DZ equations in that case, which was previously unknown.

4. Results

Polynomiality of DZ Hierarchy: Assuming Conjecture 1, the authors prove that the equations of the DZ hierarchy for any F-CohFT are differential polynomials.
Miura Equivalence: Assuming Conjecture 2, they prove that the DZ and DR hierarchies are related by a Miura transformation.
Normal Miura Equivalence: Assuming Conjecture 3, they prove the stronger result that the hierarchies are related by a normal Miura transformation (preserving the $\tau$ -structure), extending previous results from CohFTs to F-CohFTs.
Verification: The conjectures are rigorously proven for the cases $n=1$ and $g=0$ .
Reduction: Theorem 7.1 shows that the entire system of relations for $m \geq 2$ is equivalent to a finite subsystem where $\sum d_i = 2g + m - 1$ and $d_i \geq 1$ .

5. Significance

Unification of Integrable Systems: The paper provides a unified framework linking the geometry of moduli spaces (via tautological relations) to the structure of integrable hierarchies. It bridges the gap between the geometrically defined DR hierarchy and the topologically defined DZ hierarchy.
Resolution of Open Problems: It resolves the long-standing question of the polynomiality of the DZ hierarchy for general F-CohFTs and provides the first geometric formula for these equations.
New Tools: The derivation of new tautological relations using the Liu–Pandharipande localization method (specifically in the context of relative maps to $\mathbb{P}^1$ ) offers a powerful new tool for the study of the tautological ring of $\mathcal{M}_{g,n}$ .
Foundation for Future Work: By proving the conjectures for $n=1$ and $g=0$ , the authors lay the groundwork for a potential full proof for arbitrary $g$ and $n$ , which would complete the classification of the relationship between CohFTs and integrable systems.

In summary, this paper establishes a profound link between the combinatorics of stable trees, the geometry of moduli spaces, and the algebraic structure of integrable systems, offering both new conjectures and significant partial proofs that advance the understanding of the Dubrovin–Zhang and Double Ramification hierarchies.