Witt Group of Nondyadic Curves

This paper computes the derived Witt groups of smooth proper curves over nondyadic local fields of characteristic not equal to 2 by utilizing reduction techniques and conducting a general study of the existence of Theta characteristics.

The Gist

Here is an explanation of the paper "Witt Group of Nondyadic Curves" by Nanjun Yang, translated into everyday language with creative analogies.

The Big Picture: What is this paper about?

Imagine you are an architect trying to understand the structural integrity of a building. In mathematics, specifically in a field called algebraic geometry, the "building" is a curve (a shape defined by equations), and the "structural integrity" is measured by something called the Witt Group.

The Witt Group is a way of classifying "shapes" of geometric objects based on how they can be paired up or balanced. Think of it like a massive inventory system for symmetric patterns. If you can take a pattern and perfectly fold it in half so the two sides match up (a "Lagrangian"), that pattern is considered "neutral" or "zero" in this system. The Witt Group counts the patterns that cannot be folded perfectly.

The Problem:
Mathematicians have known how to calculate this inventory for curves over the real numbers (like the number line) for a long time. However, when the "ground" the curve sits on is a non-Archimedean local field (a very specific, complex type of number system used in advanced number theory, like the $p$-adic numbers), the rules get messy. Previous work only solved this for very simple curves (hyperelliptic ones).

The Solution:
Nanjun Yang has figured out how to calculate this inventory for any smooth curve over these complex number systems, provided the system isn't "dyadic" (a technical restriction meaning the number 2 behaves nicely).

---

The Core Strategy: The "Shadow" Analogy

The main trick Yang uses is reduction.

Imagine you have a complex, 3D sculpture (the curve over the complex field $K$). It's hard to analyze directly. But, if you shine a light on it, it casts a 2D shadow on the ground (the "special fiber" $X_k$ over a simpler field $k$).

Yang's method is:

1. Look at the Shadow: Analyze the simpler, singular (crumpled) shadow on the ground.
2. Reconstruct the 3D Object: Use the features of the shadow to deduce the hidden structural details of the original 3D sculpture.

The paper provides a step-by-step algorithm to translate the "cracks" and "folds" in the shadow back into the "4-torsion" (a specific type of structural weakness) of the original curve.

---

Key Concepts Explained with Metaphors

1. The "Nondyadic" Condition

Think of the number 2 as a specific type of glue. In some number systems, this glue is sticky and messy (dyadic). In others, it's clean and precise (nondyadic). Yang's paper only works when the glue is clean. This simplifies the math, allowing the "folding" rules to work predictably.

2. The "Theta Characteristic" (The Perfect Fold)

In the world of these curves, there is a special property called a Theta characteristic.

• Analogy: Imagine a piece of fabric with a pattern. A Theta characteristic exists if you can find a specific way to cut the fabric in half so that the two halves are identical mirror images.
• Why it matters: If you can find this "perfect fold," the math becomes much easier. If you can't, the structure is more complex. Yang's paper calculates exactly when this perfect fold exists based on the shape of the curve's shadow.

3. The "Bockstein Spectral Sequence" (The Filter Machine)

This sounds scary, but think of it as a multi-stage coffee filter.

• You pour a complex mixture (the raw data of the curve) into the top.
• The filter has layers. The first layer removes the big, obvious impurities. The next layer catches smaller particles.
• Yang uses this "filter machine" to separate the "4-torsion" (the specific structural flaws he cares about) from the rest of the noise. He shows that for these specific curves, the filter stops working after a few layers, making the calculation finite and manageable.

4. The "Singularities" (The Cracks in the Shadow)

When the 3D sculpture is projected onto the 2D ground, it might get crumpled. The points where it crumples are called singularities.

• Yang's algorithm looks closely at these crumpled points.
• He asks: "If I smooth out this crumple, how many new pieces do I get?"
• He counts the "odd" pieces (components with an odd number of points). This count, combined with how the pieces are connected, tells him the final answer for the Witt Group.

---

The "Recipe" for the Result

Yang's paper essentially gives a recipe to calculate the size of the Witt Group ($W(X_K)$). Here is the simplified recipe:

1. Take the Curve: Start with your smooth curve over the complex field.
2. Crush it Down: Look at its "reduction" (the shadow) over the simpler field.
3. Count the Cracks: Identify the singular points where the shadow is crumpled.
4. Check the Connections: Look at the "normalization" (the smoothed-out version of the shadow). Count how many separate pieces it has and whether they are "odd" or "even" in a specific mathematical sense.
5. The "Theta" Check: Determine if the curve has a "perfect fold" (Theta characteristic). This depends on whether certain geometric intersections are even numbers.
6. Plug into the Formula:
   • If the curve has a rational point (a specific type of anchor), the size of the Witt Group is determined by:
     • The number of "odd" pieces in the shadow.
     • The complexity of the connections between the crumpled points.
     • Whether the "perfect fold" exists.

Why is this important?

Before this paper, mathematicians were like explorers with a map that only showed the coastlines (hyperelliptic curves). They knew the shape of the simple islands but had no idea what the deep interior (general curves) looked like.

Yang has drawn the map for the entire interior. He has shown that even though the terrain is rugged and full of singularities, the "structural inventory" (the Witt Group) follows a strict, predictable pattern based on the geometry of the curve's shadow.

In summary: This paper is a masterclass in deconstruction. It takes a complex, high-dimensional mathematical object, flattens it into a simpler 2D version, analyzes the wrinkles and cracks in that flat version, and uses that information to perfectly reconstruct the hidden properties of the original object.

Technical Summary

Technical Summary: Witt Group of Nondyadic Curves

1. Problem Statement

The paper addresses the computation of the derived Witt groups $W^i(X_K)$ for smooth proper curves $X_K$ defined over nondyadic local fields $K$ (finite extensions of $\mathbb{Q}_p$ with $p>2$ or fields of the form $\mathbb{F}_{p^n}((t))$).

While the Witt groups of real algebraic curves and hyperelliptic curves with good reduction over local fields have been studied (e.g., by Knebusch, Parimala, Arason), a general computation for arbitrary smooth curves over nondyadic local fields was lacking. Specifically, previous works failed to fully characterize the 4-torsion parts of these Witt groups and the conditions under which Theta characteristics (square roots of the canonical bundle) exist in this setting.

2. Methodology

The author employs a combination of motivic homotopy theory, algebraic geometry of singular curves, and Galois cohomology. The core strategy involves reducing the problem from the generic fiber (the curve over $K$) to the special fiber (the reduction over the residue field $k$).

Key Methodological Steps:

1. Motivic Stable Homotopy Category:

   • The Witt groups are interpreted via the motivic stable homotopy category $SH(k)$.
   • The derived Witt groups $W^i(X)$ are identified with the cohomology of the Witt motivic spectrum $HW\mathbb{Z}$.
   • The author utilizes the Bockstein Spectral Sequence (BSS) associated with the exact couple of motivic cohomology. This sequence converges to the Witt motivic cohomology $H^{*,*}_W(X)$.
   • The differentials in the BSS are identified with Steenrod operations ($Sq_1, Sq_2$) and higher Bocksteins. For curves over local fields, higher Bocksteins vanish, simplifying the convergence analysis.

2. Reduction to the Special Fiber:

   • Let $X$ be a regular proper flat model of $X_K$ over the ring of integers $\mathcal{O}_K$. The special fiber $X_k$ may be singular.
   • The author analyzes the singularities of $X_k$ using the normalization $p: \tilde{X}^{red}_k \to X_k$.
   • A key tool is the study of the fundamental group and Picard schemes of the singular fiber, specifically the kernel $G$ of the map $\text{Pic}^0_{X_k/k} \to \text{Pic}^0_{\tilde{X}^{red}_k/k}$.

3. Theta Characteristics and Quadratic Forms:

   • The existence of a Theta characteristic ($\sqrt{\omega_{X_K}}$) is linked to the vanishing of the Steenrod square $Sq_2$ on the motivic cohomology group $H^{2,2}_M(X_K)$.
   • The paper establishes a correspondence between the existence of $\sqrt{\omega_{X_K}}$ and the vanishing of a specific class $\Lambda(\omega_{X_k})$ in the cohomology group $H^1(k, 2G)/R$, where $R$ is the image of a boundary map related to the Tate pairing.

4. Pairings and Invariants:

   • The author generalizes Merkurjev's pairing to singular curves to define invariants $S(X_k)$ and $G(X_k)$.
   • $S(X_k)$ relates to the cokernel of the map from 0-cycles to cohomology.
   • $G(X_k)$ captures the 2-torsion part of the fundamental group of the singular fiber.

3. Key Contributions and Results

A. General Computation Formula

The paper provides explicit formulas for the length ($l$) and the dimension of the 2-torsion part ($\dim(2W^i)$) of the derived Witt groups $W^i(X_K)$ for $i=0, 1$. The results depend on invariants of the special fiber $X_k$:

• $S(X_k)$: The cokernel of the pairing between 0-cycles and $H^1$.
• $H^2(X_k)$: Related to the Picard group of the normalization.
• $G(X_k)$: A subgroup of $H^1(k, 2G)$ related to the singularities.
• $tr(X_k)$: The number of irreducible components of the normalization with odd degree over $k$.
• $q, q', \delta$: Binary invariants determined by the existence of Theta characteristics and the behavior of the canonical bundle on the singular fiber.

Theorem 1 (Summary):
For a curve $X_K$ with a rational point:

• $l(W(X_K)) = \dim(S(X_k)) + 2\dim(H^2(X_k)) + q + 1$
• $\dim(2W(X_K))$ depends on whether $\sqrt{-1} \in K$ and involves $\dim(G(X_k))$, $tr(X_k)$, and $\delta$.
• Similar formulas are derived for $W^1(X_K)$.

B. Existence of Theta Characteristics

The paper resolves the existence of Theta characteristics ($\Theta(X_K) \neq \emptyset$) for nondyadic curves:

• $\Theta(X_K) \neq \emptyset$ if and only if $Sq_2$ vanishes on $H^{2,2}_M(X_K)$.
• This condition is reduced to the special fiber: $\Theta(X_K) \neq \emptyset$ if and only if $\Theta(X_k) \neq \emptyset$ and a specific class $\Lambda(\omega_{X_k})$ vanishes in $H^1(k, 2G)/R$.
• The paper provides a concrete algorithm to compute $\Lambda(\omega_{X_k})$ using the Tate pairing and the geometry of the singular components (Proposition 26, Definition 27).

C. Hyperelliptic Reductions

Section 8 provides a detailed algorithm for computing these invariants when the special fiber consists of hyperelliptic curves or $\mathbb{P}^1$.

• It explicitly constructs the rational functions required to determine if the canonical bundle is a square.
• It uses the Frobenius action on the roots of defining polynomials to determine the parity of self-intersections and the solvability of the associated linear systems over $\mathbb{F}_2$.

D. Elliptic Curves

Section 7 applies the general theory to elliptic curves, providing a complete classification of $W(X_K)$ based on the reduction type (Kodaira symbols $I_0, I_1, II, \dots, II^*$). The paper tabulates the values of $a$ and $b$ in the decomposition $W(X_K) \cong \mathbb{Z}/4^{\oplus a} \oplus \mathbb{Z}/2^{\oplus b}$ for all reduction types.

4. Significance

• Completeness: This work fills a major gap in the literature by providing a general algorithm for the Witt groups of curves over nondyadic local fields, moving beyond the previously known hyperelliptic or good reduction cases.
• 4-Torsion Resolution: It successfully identifies and computes the elusive 4-torsion parts of the Witt groups, which were previously unknown in this context.
• Motivic Framework: It demonstrates the power of the motivic stable homotopy category and Bockstein spectral sequences in solving classical arithmetic geometry problems, specifically linking Steenrod operations to the existence of geometric structures (Theta characteristics).
• Algorithmic Utility: The paper offers a practical, step-by-step procedure (involving normalization, Tate pairings, and specific cohomological computations) that can be applied to any smooth proper curve with a regular model, making the abstract theory computationally accessible.

In summary, Nanjun Yang's paper establishes a comprehensive framework for understanding the quadratic form theory of curves over local fields by bridging motivic homotopy theory with the arithmetic of singular reductions.