pith. sign in

arxiv: 2502.17851 · v2 · pith:YD22JMNZnew · submitted 2025-02-25 · 🧮 math.NT

Nearby cycles on the local model for the GU(n-1,1) PEL Shimura variety over a ramified prime

Joseph Muller This is my paper

Pith reviewed 2026-05-23 03:08 UTC · model grok-4.3

classification 🧮 math.NT
keywords nearby cycleslocal modelsShimura varietiesPEL typesemi-stable reductionblow-upGalois actionFrobenius eigenvalue
0
0 comments X

The pith

Nearby cycles on the local model for the GU(n-1,1) PEL Shimura variety are trivial when n is odd and reduce to one non-vanishing higher sheaf when n is even.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper computes the cohomology sheaves of the ℓ-adic nearby cycles on the local model of the PEL GU(n-1,1) Shimura variety over a ramified prime, with level the stabilizer of a self-dual lattice. It shows these sheaves are trivial for odd n. For even n only one higher cohomology sheaf survives, and the Galois action on it is given by an explicit Frobenius eigenvalue. The computation proceeds by first finding the nearby cycles on the blow-up of the isolated singularity, which has semi-stable reduction, then transferring the result to the original local model by proper base change. A reader cares because the nearby cycles control how the cohomology of the Shimura variety changes from the generic fiber to the special fiber, which in turn governs the Galois representations attached to automorphic forms.

Core claim

The cohomology sheaves of the ℓ-adic nearby cycles on the local model are trivial when n is odd; when n is even only a single higher cohomology sheaf does not vanish, and its Galois action is given by the associated Frobenius eigenvalue.

What carries the argument

Nearby cycles computed first on the blow-up at the singular point (semi-stable reduction for n ≥ 3), then transferred to the original local model via proper base change.

If this is right

  • For odd n the local model behaves as if it has good reduction with respect to nearby cycles.
  • For even n the single non-vanishing sheaf accounts for the difference between the cohomology of the special fiber and the generic fiber.
  • The explicit Frobenius eigenvalue determines the Galois representation on that cohomology sheaf.
  • The result applies to the given level structure (stabilizer of a self-dual lattice) in the GU(n-1,1) case.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the local model faithfully approximates the integral model of the Shimura variety, the same vanishing pattern would hold for the actual Shimura variety.
  • The same blow-up technique could be tested on local models for other ramified PEL types to see whether the parity of n continues to govern the vanishing.
  • The computed Frobenius eigenvalue could be checked against predictions coming from the weight-monodromy conjecture for the cohomology of the Shimura variety.

Load-bearing premise

The local model has isolated singularities and the blow-up at the singular point has semi-stable reduction.

What would settle it

For a concrete even value such as n=4, compute the nearby cycles sheaves directly on the local model and verify whether exactly one higher sheaf is non-zero and carries the predicted Frobenius eigenvalue.

Figures

Figures reproduced from arXiv: 2502.17851 by Joseph Muller.

Figure 1
Figure 1. Figure 1: The first page pEKqa,b 1 of the monodromy spectral sequence for Mψ,K [PITH_FULL_IMAGE:figures/full_fig_p014_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: The first page pEZ1qa,b 1 of the restriction of the monodromy spectral sequence to Z1 [PITH_FULL_IMAGE:figures/full_fig_p015_2.png] view at source ↗
read the original abstract

In this paper, we compute the cohomology sheaves of the $\ell$-adic nearby cycles on the local model of the PEL $\mathrm{GU}(n-1,1)$ Shimura variety over a ramified prime, with level given by the stabilizer of a self-dual lattice. This local model is known to have isolated singularities. If $n=2$ it has semi-stable reduction, and if $n\geq 3$ the blow-up at the singular point has semi-stable reduction. We compute the nearby cycles on the blow-up, then use proper base change to describe them on the original local model. As a result, we prove that the nearby cycles are trivial when $n$ is odd, and that only a single higher cohomology sheaf does not vanish when $n$ is even. In this case, we also describe the Galois action by computing the associated Frobenius eigenvalue.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper computes the ℓ-adic nearby cycles on the local model for the PEL GU(n-1,1) Shimura variety over a ramified prime, with level the stabilizer of a self-dual lattice. The local model has isolated singularities. For n=2 the model itself has semi-stable reduction; for n≥3 the blow-up at the singular point has semi-stable reduction. The authors compute the nearby cycles explicitly on the blow-up (using the semi-stable reduction to obtain the sheaves and Frobenius eigenvalues), then transfer the result to the original model via proper base change. They conclude that the nearby cycles vanish for odd n and that exactly one higher cohomology sheaf is non-vanishing for even n, with an explicit description of the associated Frobenius eigenvalue.

Significance. If the result holds, it supplies an explicit description of nearby cycles for these local models, which is useful for studying the reduction of Shimura varieties and the Galois representations on their cohomology. The vanishing statement for odd n and the single non-vanishing sheaf plus Frobenius eigenvalue for even n are concrete and potentially applicable in the Langlands correspondence. The argument relies on standard tools (proper base change, semi-stable reduction) rather than ad-hoc constructions.

major comments (2)
  1. [Introduction / statement of the strategy (following the abstract)] The claim that the blow-up yields semi-stable reduction for n≥3 (invoked to justify the explicit sheaf computation on the blow-up and the subsequent proper base change) is load-bearing for the main vanishing statements. The text states this as known but does not provide a reference or self-contained verification for the specific self-dual lattice stabilizer in the GU(n-1,1) case; if this fails to hold exactly, the transfer does not apply and the claimed vanishing for odd n (or single non-vanishing sheaf for even n) does not follow on the original model.
  2. [Section on proper base change (likely §3 or §4)] The application of proper base change to descend the nearby-cycles computation from the blow-up to the original local model requires that the relevant diagrams satisfy the étale base-change hypotheses and that the morphism is proper. The manuscript should explicitly confirm these hypotheses hold in the present setting rather than treating them as automatic.
minor comments (2)
  1. The abstract is concise; the introduction would benefit from a numbered statement of the main theorem.
  2. Ensure that all prior references establishing isolated singularities of the local model are cited explicitly.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their detailed reading and for highlighting two points that require additional clarification in the manuscript. Both comments concern the justification of key steps in our strategy (semi-stable reduction of the blow-up and the hypotheses for proper base change). We address each below and will revise the text accordingly to strengthen the exposition.

read point-by-point responses

  1. Referee: [Introduction / statement of the strategy] The claim that the blow-up yields semi-stable reduction for n≥3 (invoked to justify the explicit sheaf computation on the blow-up and the subsequent proper base change) is load-bearing for the main vanishing statements. The text states this as known but does not provide a reference or self-contained verification for the specific self-dual lattice stabilizer in the GU(n-1,1) case; if this fails to hold exactly, the transfer does not apply and the claimed vanishing for odd n (or single non-vanishing sheaf for even n) does not follow on the original model.

    Authors: We agree that the semi-stable reduction property of the blow-up is central to the argument and that the manuscript should make the justification fully explicit rather than treating it as standard background. In the revised version we will add a precise reference to the literature establishing semi-stable reduction for these local models (in particular, results on local models for unitary groups with parahoric level at ramified primes) together with a short paragraph confirming that the property holds for the self-dual lattice stabilizer in the GU(n-1,1) setting. If a fully self-contained verification is preferred, we can also sketch the local computation of the special fiber after blow-up. revision: yes

  2. Referee: [Section on proper base change (likely §3 or §4)] The application of proper base change to descend the nearby-cycles computation from the blow-up to the original local model requires that the relevant diagrams satisfy the étale base-change hypotheses and that the morphism is proper. The manuscript should explicitly confirm these hypotheses hold in the present setting rather than treating them as automatic.

    Authors: We accept that the hypotheses for proper base change should be stated explicitly. In the revision we will insert a short subsection (or paragraph) verifying that the blow-up morphism is proper, that the relevant diagrams are Cartesian in the étale topology, and that the base-change theorem for nearby cycles therefore applies without additional conditions in this case. This will make the transfer from the blow-up to the original local model fully rigorous. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation relies on external theorems and cited properties of the local model.

full rationale

The paper states that the local model has isolated singularities (known) and that the blow-up has semi-stable reduction for n≥3 (invoked as a fact). It then computes nearby cycles on the blow-up and transfers via the standard proper base change theorem. These steps cite external results rather than deriving the target vanishing statements from the result itself or from self-citations that reduce the claim. No self-definitional equations, fitted predictions, or load-bearing self-citation chains appear in the provided abstract and description. The central claims about triviality for odd n and single non-vanishing sheaf for even n follow from the explicit computation on the resolved model, which is independent of the final output.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The computation rests on two domain facts about the local model that are stated as known rather than derived in the paper.

axioms (2)
  • domain assumption The local model has isolated singularities.
    Invoked to justify the blow-up step and the subsequent base-change argument.
  • domain assumption For n ≥ 3 the blow-up at the singular point has semi-stable reduction.
    Used to compute nearby cycles on the resolved model before transferring back.

pith-pipeline@v0.9.0 · 5691 in / 1508 out tokens · 32192 ms · 2026-05-23T03:08:34.578877+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

4 extracted references · 4 canonical work pages

  1. [1]

    Expos´ e I : Autour du th´ eor` eme de monodromie lo cale

    [Ill94] L. Illusie. “Expos´ e I : Autour du th´ eor` eme de monodromie lo cale”. In: P´ eriodesp-adiques - S´ eminaire de Bures, 1988 . Ed. by Jean-Marc Fontaine. Ast´ erisque

    work page 1988

  2. [2]

    Local Models for Ramified Unitary Groups

    [Kr¨ a03] N. Kr¨ amer. “Local Models for Ramified Unitary Groups”.In: Abh.Math.Semin.Univ.Hambg. 73.1 (2003), pp. 67–80. doi: 10.1007/BF02941269. [Mil13] J. S. Milne. Lectures on Etale Cohomology (v2.21). Available at www.jmilne.org/math/

    work page doi:10.1007/bf02941269 2003

  3. [3]

    On the arithmetic moduli schemes of PEL Shimur a vari- eties

    [Pap00] G. Pappas. “On the arithmetic moduli schemes of PEL Shimur a vari- eties”. In: Journal of Algebraic Geometry 9 (2000). [RZ96] M. Rapoport and T. Zink. Period Spaces for ”p”-divisible Groups (AM- 141). Princeton University Press,

    work page 2000

  4. [4]

    Weight Spectral Sequence and Independence of l

    isbn: 9780691027814. [Sai03] T. Saito. “Weight Spectral Sequence and Independence of l”. In: Journal of the Institute of Mathematics of Jussieu 2.4 (2003), pp. 583–634. doi: 10.1017/S1474748003000173. [Shi22] Y Shi. “Special Cycles on Unitary Shimura Curves at Ramified Pr imes”. In: manuscripta math. (2022). doi: 10.1007/s00229-022-01412-z . [Wei49] A. Wei...

    work page doi:10.1017/s1474748003000173 2003