TL;DR
This paper establishes a powerful bridge between algebraic geometry and topology by proving that the Grothendieck-Witt and Witt groups of smooth complex cellular varieties are isomorphic to their topological real K-theory (KO-groups). By leveraging $\mathbb{A}^1$-homotopy theory, the author provides a unified framework to compute these complex algebraic invariants using well-established topological tools, specifically calculating values for quadrics, spinor varieties, and Grassmannians.
What Problem Does This Paper Address?
In algebraic geometry, Witt groups are invariants that classify quadratic forms (symmetric bilinear forms) over a variety. While they are fundamental to understanding the geometry of a space, they are notoriously difficult to compute. Unlike standard algebraic K-theory ($K_0$), which behaves predictably for "cellular" varieties (spaces built out of simple affine cells), Witt groups are "non-oriented." This means they don't decompose as easily into the sum of their parts, often leaving researchers guessing how many copies of $\mathbb{Z}/2$ or other groups to expect.
The paper tackles the challenge of finding a systematic, topological way to calculate these algebraic groups for a broad class of spaces: smooth complex cellular varieties. This category includes many important geometric objects like projective spaces, Grassmannians, and flag varieties.
Methods
The author utilizes the framework of $\mathbb{A}^1$-homotopy theory (developed by Morel and Voevodsky) to bridge the gap between the algebraic and topological worlds.
The core of the method involves:
- Representability: Relying on the fact that hermitian K-theory (the theory underlying Witt groups) can be represented by a spectrum in the stable $\mathbb{A}^1$-homotopy category.
- Comparison Maps: Defining a "complex realization" functor that maps algebraic varieties to their underlying topological spaces and algebraic spectra to topological spectra (like the KO-spectrum).
- Cellular Induction: Proving the isomorphism by induction on the number of cells. The author shows that the comparison maps respect the "gluing" instructions (localization sequences) used to build complex varieties.
- Atiyah-Hirzebruch Spectral Sequence (AHSS): Once the problem is moved into the topological realm, the author uses the AHSS—a standard tool in algebraic topology—to perform the actual numerical computations for specific varieties.
Datasets
Not applicable. This is a theoretical mathematics paper; it does not use empirical datasets or benchmarks.
Key Insights and Ideas
- The Symmetric Link: A key geometric intuition is that symmetric complex vector bundles (the building blocks of Witt groups) are in one-to-one correspondence with real vector bundles in topology. This is why topological KO-theory (K-theory of real bundles) is the natural counterpart to algebraic Witt groups.
- $\mathbb{A}^1$-Homotopy as a Glue: The use of $\mathbb{A}^1$-homotopy theory is the "aha moment" that allows the author to prove the comparison maps are compatible with the long exact sequences needed for induction. Without this framework, proving that the algebraic and topological "gluing" matches up is much harder.
- Handling "Twists": The paper successfully extends the results to "twisted" Witt groups—those where the quadratic forms take values in a line bundle rather than the structure sheaf. This is crucial because many geometric varieties (like certain quadrics) naturally involve these twists.
Key Results
The main result is Theorem 2.5, which identifies:
- The Grothendieck-Witt group $GW^q(X)$ with the topological group $KO^{2q}(X(\mathbb{C}))$.
- The Witt group $W^q(X)$ with the topological group $KO^{2q-1}(X(\mathbb{C}))$.
Using this theorem, the author provides explicit tables for the Witt groups of:
- Projective Spaces: Recovering and unifying previous results by Arason and Walter.
- Grassmannians: Providing an alternative topological computation for both untwisted and twisted cases.
- Smooth Complex Quadrics: Completing calculations that had previously been partially unpublished or handled through more fragmented methods.
- Spinor Varieties and Exceptional Hermitian Symmetric Spaces: Extending the reach of Witt group computations to these more complex homogeneous spaces.
Limitations and Future Work
- The Cellular Constraint: The results are strictly for cellular varieties. While this covers many important cases, it excludes varieties with more complex "non-rational" cohomology.
- Ground Field: The paper focuses on varieties over the complex numbers ($\mathbb{C}$). Extending these results to varieties over real numbers ($\mathbb{R}$) or finite fields would involve significantly different topological realizations (e.g., involving $C_2$-equivariant topology).
- Foundational Assumptions: At the time of writing, the author relied on certain "standing assumptions" regarding the representability of hermitian K-theory in $\mathbb{A}^1$-homotopy theory that were then only available in draft form (specifically by Morel). While these are now widely accepted, they represented a dependency on cutting-edge foundational work at the time.
Summary updated