Stacky geometry and logarithmic topology of transversely affine foliations

arXiv:2605.27669 · math.AG, math.CV, math.DG · Submitted 2026-05-26 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Mathematics. The day in mathematics: new theorems, new proofs, and what they open.

Ada: I'm Ada.

Emmy: And I'm Emmy, here with Kolya.

Ada: We're looking at the paper: "Stacky geometry and logarithmic topology of transversely affine foliations".

Emmy: Here's the gist of the paper. We study the stacky and logarithmic-topological structures associated with transversely affine foliations to provide a geometric counterpart for Singer's classical theorem on Liouvillian first integrals The quotient stack controls the linear part of the theory, while the Kato–Nakayama space captures its logarithmic-topological and dynamical content.

Ada: Geometric Foundation of Transverse Structures The paper introduces transversely affine structures as flat meromorphic connections on the normal bundle, where a singular transverse affine structure is defined as a flat meromorphic connection along a reduced divisor D such that the zero locus has codimension at least 2 On an open set U where the normal bundle is trivial, this structure is given by equations dω0 = ω0 ∧ η0 and dη0 = 0, which are related to a gauge relation.

Emmy: The affine monodromy representation ξ: π1(U) → Aff(C) has a linear part ρ: π1(U) → C∗, and the quotient stack

C∗/Γ: is the natural target for the multiplicative developing coordinate, recording the linear monodromy.

Ada: Logarithmic Topology and Boundary Dynamics The Kato–Nakayama space Xlog is locally a real oriented blow-up of the divisor, with torus fibres over the boundary strata.

Paper summary: Ada: So to summarize this paper, Barassa et al. are studying stacky geometry and logarithmic topology associated with transversely affine foliations >

Emmy: They claim that the quotient stack gives us the natural geometric base for the multiplicative developing coordinate, which is how you track things linearly >

Ada: And they also look at the Kato–Nakayama space, which is their way of capturing that logarithmic-topological and dynamical content of the theory >

Kolya: It sounds like they're trying to bridge the gap between purely algebraic stacky pictures and something that has a boundary or topological feel attached to it >

Ada: That’s right. They show that every holomorphic stacky first integral for the same foliation is obtained through post-composition with an endomorphism of this base stack >

Emmy: They also enlarge this picture in Theorem four point five, where things like Kummer or root-stack refinements are interpreted as the geometric analogue of Singer’s algebraic extension step > <ref:2605.27669#pg2,Kummer or root-stack refinements are interpreted as the geometric>

Conclusion: Ada: So to wrap up, the title itself points to how they are using stacky geometry—that’s the quotient stack—alongside logarithmic topology to understand these foliations >

Emmy: And the authors are connecting Singer’s classical theorem on first integrals with this new geometric framework by showing how different refinements of their structures correspond to algebraic steps in the theory >

Ada: The big implication is that they give us a concrete way to understand the linear behavior of these complex foliations through this quotient stack, while the logarithmic space gives us that necessary boundary and dynamical information >

Kolya: For someone just listening, it means that instead of just seeing an algebraic relation, you get a picture where you can see both the linear structure and how those structures behave near some boundaries or special points >

Coordenaตo de Aperfeiตamento de Pessoal de Nivel Superior - Brasil (CAPES) · CNPq · University of Bari · INdAM-GNSAGA

math.AG, math.CV, math.DG

Submitted: 2026-05-26

Updated: 2026-11-09

Comments: 26 pages

License: http://creativecommons.org/licenses/by/4.0/

Importance score: 92/100

The gist: We study the stacky and logarithmic-topological structures associated with transversely affine foliations to provide a geometric counterpart for Singer's classical theorem on Liouvillian first…

Key concepts

Transversely Affine Structures
These are flat meromorphic connections on the normal bundle of a foliation. A singular version occurs when these connections have singularities along a reduced divisor D, allowing for analysis of complex geometric structures related to the foliation's transverse geometry.
Quotient Stack [C*/Γ]
This stack represents the linear part of the theory, controlling how monodromy acts on coordinates. It is the natural target for developing multiplicative coordinates and records the linear monodromy associated with transversely affine structures.
Kato–Nakayama Space (Xlog)
This space provides the logarithmic-topological counterpart to the stacky picture. It is locally a blow-up of a divisor where torus fibers over boundary strata, capturing the essential dynamical content and boundary behavior of the foliation.

Terminology

Summary

We study the stacky and logarithmic-topological structures associated with transversely affine foliations to provide a geometric counterpart for Singer's classical theorem on Liouvillian first integrals The quotient stack controls the linear part of the theory, while the Kato–Nakayama space captures its logarithmic-topological and dynamical content.

Geometric Foundation of Transverse Structures

The paper introduces transversely affine structures as flat meromorphic connections on the normal bundle, where a singular transverse affine structure is defined as a flat meromorphic connection along a reduced divisor D such that the zero locus has codimension at least 2 On an open set U where the normal bundle is trivial, this structure is given by equations dω0 = ω0 ∧ η0 and dη0 = 0, which are related to a gauge relation. The affine monodromy representation ξ: π1(U) → Aff(C) has a linear part ρ: π1(U) → C∗, and the quotient stack [C∗/Γ] is the natural target for the multiplicative developing coordinate, recording the linear monodromy. p. 0

Where the paper says it

Their analytic continuation is governed by the affine monodromy representation ξ : π1(U) → Aff(C) = C∗ ⋉ C, whose linear part is the character ρ : π1(U) → C∗ . Writing Γ = im(ρ), the quotient stack [C∗/Γ ] is the natural target for the multiplicative developing coordinate: it records the linear monodromy and organises the residual reparametrisation freedom.

Logarithmic Topology and Boundary Dynamics

The Kato–Nakayama space Xlog is locally a real oriented blow-up of the divisor, with torus fibres over the boundary strata. On these fibres, the residues of η define canonical boundary characters, while the pull-back π∗η induces a real codimension-two foliation on Xlog itself. p. 0

Where the paper says it

The Kato–Nakayama space X log is locally a real oriented blow-up of the divisor, with torus fibres over the boundary strata; see Definition 7. On these fibres, the residues of η define canonical boundary characters, while the pull-back π ∗η induces a real codimension-two foliation on X log itself.

The Kato–Nakayama space furnishes the logarithmic-topological counterpart to the stacky picture. p. 1
Where the paper says it

CORRÊA attracting, repelling, and purely angular boundary components. In this way, the Kato–Nakayama space furnishes the logarithmic-topological counterpart to the stacky picture: Kummer refinements remove torsion in the meridional monodromy, whereas X log retains the angular and dynamical content of the same residue data.

Classification of Reparametrizations

The classification of reparametrisations is carried out regime by regime based on the geometry of the holonomy group Γ. In non-discrete regimes, semi-equivariance is rigid, meaning holomorphic and meromorphic endomorphisms are necessarily monomial. p. 1

Where the paper says it

The structure of finitely generated subgroups of C∗ and the corresponding proper or non-proper dichotomy are analysed in Propositions 5. In the non-discrete regimes, semi-equivariance is rigid: holomorphic and meromorphic endomorphisms are necessarily monomial, as shown in Theorems 5.

p. 1
Where the paper says it

The structure of finitely generated subgroups of C∗ and the corresponding proper or non-proper dichotomy are analysed in Propositions 5. In the non-discrete regimes, semi-equivariance is rigid: holomorphic and meromorphic endomorphisms are necessarily monomial, as shown in Theorems 5.

In torsion and elliptic regimes, descriptions are obtained in specific propositions. p. 1
Where the paper says it

In the non-discrete regimes, semi-equivariance is rigid: holomorphic and meromorphic endomorphisms are necessarily monomial, as shown in Theorems 5. In the torsion and elliptic regimes, the corresponding descriptions are obtained in Propositions 5.

The classification of endomorphisms is summarized by the structure of finitely generated subgroups of C∗. p. 1
Where the paper says it

The classification of the resulting reparametrisations is then carried out regime by regime. The structure of finitely generated subgroups of C∗ and the corresponding proper or non-proper dichotomy are analysed in Propositions 5.

p. 1
Where the paper says it

The classification of the resulting reparametrisations is then carried out regime by regime. The structure of finitely generated subgroups of C∗ and the corresponding proper or non-proper dichotomy are analysed in Propositions 5.

Synthesis via Singer Operations

The paper proves that every holomorphic stacky first integral for the same transversely affine foliation is obtained, up to 2-isomorphism, by post-composition of the canonical one with an endomorphism of [C∗/Γ]. This picture is enlarged in Theorem 4. p. 1

Where the paper says it

15 that every holomorphic stacky first integral for the same transversely affine foliation is obtained, up to 2-isomorphism, by post-composition of F with an endomorphism of the stacky base. This picture is enlarged in Theorem 4.

p. 1
Where the paper says it

15 that every holomorphic stacky first integral for the same transversely affine foliation is obtained, up to 2-isomorphism, by post-composition of F with an endomorphism of the stacky base. This picture is enlarged in Theorem 4.

5, where Kummer or root-stack refinements are interpreted as the geometric counterpart of Singer’s algebraic-extension step. p. 1
Where the paper says it

This picture is enlarged in Theorem 4. 5, where Kummer, or root-stack, refinements are interpreted as the geometric counterpart of Singer’s algebraic-extension step.

p. 1
Where the paper says it

This picture is enlarged in Theorem 4. 5, where Kummer, or root-stack, refinements are interpreted as the geometric counterpart of Singer’s algebraic-extension step.

The operations exhaust the geometric analogues of Singer’s closure steps, once the holonomy constants are fixed. p. 0
Where the paper says it

INTRODUCTION Singer’s classical Theorem on Liouvillian first integrals points to a strong rigidity phenomenon in the transversely affine setting: the relevant transcendental operations are generated by quadratures, exponentials of primitives, and algebraic extensions [29]. In the present geometric setting, it is natural to recast this pattern in terms of closed logarithmic 1-forms, monodromy representations, and developing maps [4, 13].

Boundary Realizations and Complementary Structures

The same residue data admit two complementary boundary realisations: a Kummer one, which removes torsion in the meridional monodromy, and a Kato–Nakayama one, which retains the angular boundary data. The quotient stack [C∗/Γ] controls the linear and reparametrisation-theoretic aspect of the theory, while the Kato–Nakayama space provides the natural topological and dynamical setting in which the boundary geometry of transversely affine foliations becomes visible. p. 1

Where the paper says it

15 show that the same holonomy group Γ governs both the transverse stacky quotient and the logarithmic-topological boundary model. The quotient stack [C∗/Γ ] controls the linear and reparametrisation-theoretic aspect of the theory, while the Kato– Nakayama space provides the natural topological and dynamical setting in which the boundary geometry of transversely affine foliations becomes visible.

Automorphism Groups of Bases

The automorphism groups are distinguished into two layers: π0 Aut2(X) governs the classification of reparametrizations, while π1 Aut2(X) measures the intrinsic isotropy carried by the stack itself. The open base behaves like an honest curve, whereas the compactified base carries nontrivial 2-categorical information concentrated at its stacky boundary points. p. 1

Where the paper says it

Finally, Theorems 6. 3 compute the automorphism group of the open stacky base and the compactified automorphism 2-group.

Boundary Dynamics and Leaf Geometry

The boundary characters constructed above constitute the first layer of the logarithmic boundary structure attached to η. The induced real codimension-two foliation is integrable, and its leaves are governed entirely by the arithmetic of these residues. p. 0

Where the paper says it

On these fibres, the residues of η define canonical boundary characters, while the pull-back π ∗η induces a real codimension-two foliation on X log itself. The restriction of this foliation to the boundary tori is linear in the angular variables, and its behaviour is controlled by the arithmetic of the residue vector: according to the rational relations among the residues, one obtains closed leaves, Kronecker-type dense leaves, or leaves dense in a proper subtorus.

The final statement is that the boundary dynamics is governed entirely by the arithmetic of these residues, more precisely by the rational relations they satisfy. p. 0
Where the paper says it

On these fibres, the residues of η define canonical boundary characters, while the pull-back π ∗η induces a real codimension-two foliation on X log itself. The restriction of this foliation to the boundary tori is linear in the angular variables, and its behaviour is controlled by the arithmetic of the residue vector: according to the rational relations among the residues, one obtains closed leaves, Kronecker-type dense leaves, or leaves dense in a proper subtorus.

The Logarithmic Lift and Target Space

The interior developing map admits a canonical logarithmic lift Flog: Xlog → [C∗/Γ]. This lift is constructed using the local logarithmic formula zlogW defined in (8. p. 1

Where the paper says it

These two sides meet in the logarithmic lift of the developing map. Using the explicit local formula of Lemma 8.

1). The compactified target records the radial limits of the logarithmic lift, but not the full boundary behaviour. p. 1
Where the paper says it

Its compactified radial behaviour is described in Corollary 8. More generally, the boundary analysis carried out in Propositions 8.

The Role of Kummer Refinements

If residue exponentials are torsion, Kummer refinement along 0 and ∞ realizes the algebraic extension step in Singer’s closure process. This alters the base by replacing Γ with µN × Γ, but does not give rise to any new analytic reparametrizations. p. 0

Where the paper says it

We then classify these reparametrisations according to the geometry of the holonomy group. On the logarithmic-topological side, we pass to the Kato–Nakayama space, where the residues define canonical boundary characters, govern the induced linear dynamics on the boundary tori, and give rise to a canonical logarithmic lift of the developing map.

The Role of Kato–Nakayama Space

The Kato–Nakayama space records the angular part of the same residue data, which is not visible at the level of a purely radial compactification. The final statement is therefore just the comparison of these two mechanisms: Kummer base change kills torsion in the meridional monodromy, whereas the latter retains the angular boundary information determined by the same residues. p. 1

Where the paper says it

CORRÊA attracting, repelling, and purely angular boundary components. In this way, the Kato–Nakayama space furnishes the logarithmic-topological counterpart to the stacky picture: Kummer refinements remove torsion in the meridional monodromy, whereas X log retains the angular and dynamical content of the same residue data.

Final Synthesis of Boundary Realizations

The theorem shows that the same residue data admit two complementary boundary realisations: a Kummer one, which trivialises the torsion in the meridional monodromy, and a Kato–Nakayama one, which retains the angular boundary data. The final statement is therefore just the comparison of these two mechanisms. p. 9

Where the paper says it

In general no such map exists, because the translational part of H does not preserve the multiplicative structure of C∗ . Thus the natural comparison between the two theories is through the common base BΓ .

p. 9
Where the paper says it

In general no such map exists, because the translational part of H does not preserve the multiplicative structure of C∗ . Thus the natural comparison between the two theories is through the common base BΓ .

The Structure of Automorphisms

Aut1([C∗/Γ]) =∼ (C∗/Γ) ⋊ (Z/2), where the nontrivial element of Z/2 acts on C∗/Γ by inversion. The induced automorphism of Γ is therefore only the discrete ±1 part. p. 5

Where the paper says it

This proves the stated description of End(Γ ). The statement for automorphisms is immediate: a class in End(Zk+m, L) Hom(Zk+m, L) defines an automorphism of Γ if and only if the induced endomorphism of Zk+m/L =∼ Γ is invertible.

The Compactified Base Automorphisms

Aut(P1) = PGL2(C), and there is no nontrivial π1. The inertia at 0 and ∞ is nontrivial, so the automorphism object is genuinely 2-categorical. p. 6

Where the paper says it

an assignment to each object z ∈ C∗ of an arrow z → z, hence an element of the stabilizer at z. Since the action on C∗ is free, stabilizers are trivial, and the only freedom comes from changing the strict lift by a right Γ -equivariant automorphism of the pulled-back torsor, i.

Leaf Geometry on Boundary Torus

The leaves on the boundary torus are diffeomorphic to L =∼ V/(V ∩ ΛI) =∼ Rl−rI−sI × (S1)sI. The final statement is the content of Proposition 7. p. 1

Where the paper says it

In this way, the Kato–Nakayama space furnishes the logarithmic-topological counterpart to the stacky picture: Kummer refinements remove torsion in the meridional monodromy, whereas X log retains the angular and dynamical content of the same residue data. We prove in Proposition 2.

8.

Boundary Dynamics Governed by Residues

The boundary dynamics is governed entirely by the arithmetic of these residues, more precisely by the rational relations they satisfy. The final statement is therefore just the comparison of these two mechanisms.

Final Theorem on Boundary Realizations

The logarithmic boundary of the transversely affine structure is canonically realised on the Kato–Nakayama space Xlog in the following sense: (i) for every boundary stratum D◦I and every point x ∈ D◦I, the residues of η determine a boundary character χI,x: π1(π−1(x)) → C∗; (ii) the pull-back π∗η induces on π−1(U) ⊂ Xlog a canonical real foliation of codimension two, whose restriction to the boundary tori is linear in the angular variables and whose topology and closures are governed by the arithmetic of the residue vector; (iii) the interior developing map admits a canonical logarithmic lift Flog: Xlog → [C∗/Γ], Γ = im(ρ). The final statement is exactly Theorem 8. p. 4

Where the paper says it

Likewise, hj = ρ(δj) = exp Z δj η ! This proves the final assertion.

15.

References

[1] K Behrend and B Noohi, Uniformization of Deligne–Mumford curves, J Reine Angew Math. 599 (2006), 111–153

[2] K Behrend and P Xu, Differentiable stacks and gerbes, J Symplectic Geom. 9 (2011), no. 3, 285–341

[3] C Cadman, Using stacks to impose tangency conditions on curves, Amer J Math. 129 (2007), no. 2, 405–427

[4] G Cousin and J V Pereira, Transversely affine foliations on projective manifolds, Math Res Lett. 21 (2014), no. 5, 985–1014

[5] D Carchedi, S Scherotzke, N Sibilla, and M Talpo, Kato–Nakayama spaces, infinite root stacks and the profinite homotopy type of log schemes, Geom. Topol. 21 (2017), no. 5, 3093–3158

[6] M Corrêa, P Perrella and S Velazquez, The logarithmic leaf complex and foliated d-semistability, arXiv:2605.00285v1, 2026

[7] P Deligne, Équations différentielles à points singuliers réguliers, Lecture Notes in Math., vol. 163, Springer, 1970

[8] R M Goresky, Triangulation of stratified objects, Proc Amer Math Soc. 72 (1978), no. 1, 193–200

[9] J Hall, A relative GAGA principle for families of curves, J Lond Math Soc. (2) 90 (2014), no. 1, 29–48

[10] F E A Johnson, On the triangulation of stratified sets and singular varieties, Trans Amer Math Soc. 275 (1983), no. 1, 333–343

[11] K Kato and C Nakayama, Log Betti cohomology, log étale cohomology, and log de Rham cohomology of log schemes over C, Kodai Math. J. 22 (1999), no. 2, 161–186

[12] E Lerman, Orbifolds as stacks?, Enseign. Math. (2) 56 (2010), no. 3–4, 315–363

[13] F Loray, J V Pereira, and F Touzet, Representations of quasi-projective groups, flat connections and transversely projective foliations, J Éc. polytech. Math. 3 (2016), 263–308

[14] M McQuillan, Canonical models of foliations, Pure Appl Math Q. 4 (2008), no. 3, 877–1012

[15] I Moerdijk and D A Pronk, Orbifolds, sheaves and groupoids, K-Theory 12 (1997), no. 1, 3–21

[16] B Noohi, Fundamental groups of algebraic stacks, J Inst Math Jussieu 3 (2004), no. 1, 69–103

[17] B Noohi, Foundations of topological stacks. I, arXiv:math/0503247

[18] B Noohi, Homotopy types of topological stacks, Adv Math. 230 (2012), no. 4–6, 2014–2047

[19] B Noohi, Group actions on algebraic stacks via butterflies, J of Algebra. 486 (2017), 36–63

[20] B Noohi, Automorphism 2-group of a weighted projective stack, MPIM Preprint Ser. (2007), no. 56

[21] A Ogus, On the logarithmic Riemann–Hilbert correspondence, Doc Math Extra Vol. (2003), 655–724

[22] M Olsson, Algebraic Spaces and Stacks, Amer Math Soc Colloq Publ., vol. 62, American Mathematical Society, 2016

[30] M Talpo and A Vistoli, The Kato–Nakayama space as a transcendental root stack, Int Math Res Not IMRN 2018, no. 19, 6145–6176

BARE

The single most informative sentence about the paper's subject and finding is The quotient stack controls the linear part of the theory, while the Kato–Nakayama space captures its logarithmic-topological and dynamical content.

The classification of reparametrisations is carried out regime by regime based on the geometry of the holonomy group Γ. In non-discrete regimes, semi-equivariance is rigid, meaning holomorphic and meromorphic endomorphisms are necessarily monomial. The classification of endomorphisms is summarized by the structure of finitely generated subgroups of C∗.

The paper proves that every holomorphic stacky first integral for the same transversely affine foliation is obtained, up to 2-isomorphism, by post-composition of the canonical one with an endomorphism of [C∗/Γ]. This picture is enlarged in Theorem 4.5, where Kummer or root-stack refinements are interpreted as the geometric counterpart of Singer’s algebraic-extension step.

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved AI systems can perform geometric analysis of transversely affine foliations by constructing canonical logarithmic lifts F log X log to quotient stacks [C∗/Γ] and compactified targets [P1/Γ].

  2. The system can classify reparametrizations by identifying them with endomorphisms of the stacky base [C∗/Γ] and analyzing these endomorphisms based on the geometry of the holonomy group Γ.

  3. The AI can perform logarithmic-topological analysis on the Kato–Nakayama space X log, using residue data to define canonical boundary characters and analyze induced linear dynamics on boundary tori.

  4. The system can distinguish between two complementary boundary realisations: a Kummer one that trivialises torsion in the meridional monodromy and a Kato–Nakayama one that retains angular boundary data.

  5. The AI can classify the automorphism group of the open base [C∗/Γ] as (C∗/Γ) ⋊ (Z/2) and the automorphism 2-group of the compactified base [P1/Γ] as Γ x Γ, with a specified action.

  6. The system can perform regime-by-regime classification of Singer operations by relating them to Kummer refinements and geometric analogues of algebraic extensions via morphisms between one-dimensional stacky transverse bases.

Abstract

We study the linear monodromy of codimension-one transversely affine foliations through multiplicative quotient stacks and logarithmic boundary topology. When the transverse affine structure admits a global presentation on U, a choice of multiplicative developing coordinate induces a morphism to [C*/Γ], where Γ C* is the image of the linear monodromy. We prove that for infinite Γ every strict semi-equivariant holomorphic self-map of C* is a monomial, and classify the nonconstant strict semi-equivariant rational self-maps of P 1 for arbitrary Γ. For a logarithmic connection along a simple normal crossings divisor, we describe the meridional characters and angular foliations on the boundary fibres of the Kato--Nakayama space, including the topology and closures of their leaves. We also describe the extension of the linear monodromy local system to this space and the effect of local Kummer base change on finite meridional monodromy.

Sources

Related papers