[Paper Review] On the middle convolution
This paper provides a cohomological interpretation of the algebraic middle convolution functor $MC_{ heta}$ on local systems over the punctured complex line, establishing a Riemann-Hilbert correspondence via higher direct image sheaves. It proves that $MC_{ heta}(\mathcal{F}) \cong R^1\bar{p}_2_* (j_* (\mathrm{pr}_1^*\mathcal{F} \otimes q^*\mathcal{L}_\theta))$, enabling an algorithmic construction of Fuchsian systems for rigid local systems and verifying the Grothendieck-Katz $p$-curvature conjecture for new classes of differential equations.
We present a cohomological interpretation of the middle convolution functor MC and find an explicit Riemann-Hilbert correspondence for MC_λ. This leads to an algorithm for the construction of Fuchsian systems which correspond to irreducible rigid local systems under the Riemann-Hilbert correspondence. Also, the effect of MC_λon the p-curvatures ist determined and new examples of differential systems are found which satisfy the Grothendieck-Katz p-curvature conjecture.
Motivation & Objective
- To provide a cohomological interpretation of the algebraic middle convolution functor $MC_{\lambda}$, mirroring Katz’s $l$-adic middle convolution.
- To establish a Riemann-Hilbert correspondence for $MC_{\lambda}$, translating it into a differential system with regular singularities.
- To develop an algorithm for constructing Fuchsian systems corresponding to irreducible rigid local systems.
- To analyze the behavior of $MC_{\lambda}$ on $p$-curvatures and verify the Grothendieck-Katz $p$-curvature conjecture for new examples.
Proposed method
- Use of singular cohomology and sheaf-theoretic constructions on the fibration $p_2: E \to X$, where $E = \{(x,y) \in \mathbb{C}^2 \mid x,y \ne t_i, x \ne y\}$.
- Application of the higher direct image sheaf $R^1\bar{p}_2_* (j_* (\mathrm{pr}_1^*\mathcal{F} \otimes q^*\mathcal{L}_\lambda))$ to realize $MC_\lambda(\mathcal{F})$.
- Employment of Pochhammer contours and twisted bases to model monodromy and convolution via crossed homomorphisms.
- Construction of differential systems in Okubo normal form to analyze the effect of $MC_\lambda$ on the system's structure.
- Use of gauge transformations and block matrix decompositions to simplify the system and analyze nilpotency.
- Application of the Leray spectral sequence and motivic Galois group theory to verify the $p$-curvature conjecture for certain systems.
Experimental results
Research questions
- RQ1How can the algebraic middle convolution $MC_\lambda$ be interpreted cohomologically in the complex setting?
- RQ2What is the explicit Riemann-Hilbert correspondence for $MC_\lambda$ on local systems over $\mathbb{C} \setminus \{t_1,\dots,t_r\}$?
- RQ3Can $MC_\lambda$ be used to algorithmically construct Fuchsian systems for irreducible rigid local systems?
- RQ4What is the behavior of $MC_\lambda$ on $p$-curvatures, and does it preserve the Grothendieck-Katz $p$-curvature conjecture?
- RQ5Under what conditions is the convolution of a globally nilpotent system also globally nilpotent?
Key findings
- The middle convolution $MC_\lambda(\mathcal{F})$ is isomorphic to the higher direct image sheaf $R^1\bar{p}_2_* (j_* (\mathrm{pr}_1^*\mathcal{F} \otimes q^*\mathcal{L}_\lambda))$, providing a cohomological realization of the functor.
- The construction yields an algorithm to produce Fuchsian systems with regular singularities whose solution sheaf is $MC_\lambda(\mathcal{F})$ for any given rigid local system $\mathcal{F}$.
- For $\mu = -1$, the convolution $c_{-1}(a(\mathfrak{p}))^{k+1} = 0$ and $mc_{-1}(a(\mathfrak{p}))^{k+1} = 0$, showing nilpotency in the mod $p$ reduction.
- When $\mu = n_1/n_2$ with $p \nmid n_1n_2$, it holds that $c_{\mu-1}(a(\mathfrak{p}))^{k+2} = 0$ and $mc_{\mu-1}(a(\mathfrak{p}))^{k+2} = 0$, confirming nilpotency under these conditions.
- The Grothendieck-Katz $p$-curvature conjecture holds for the system $D(L_n, \mathbf{a}, \mu)$, as shown via the connectivity of the motivic Galois group of the fiber.
- The system $D(L_n, \mathbf{a}, \mu)$ is globally nilpotent of rank 3, as established by Theorem 7.9 and Lemma 7.11.
Better researchstarts right now
From reading papers to final review, dramatically reduce your research time.
No credit card · Free plan available
This review was created by AI and reviewed by human editors.