Trigonal Canonically Fibered Surfaces

Houari Benammar Ammar; Nathan Grieve; Xi Chen

REVIEW 2 major objections 4 minor 10 references

Trigonal Canonically Fibered Surfaces

T0 review · 2 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A new proof settles Xiao's genus-5 boundedness conjecture for canonically fibered surfaces.

desk verdict Substantial fix to Xiao's conjecture for trigonal genus 5, with a clean bound pg<=50, but the non-trigonal reduction still leans on an unproved lemma from the very paper it claims to replace. read the letter →

arxiv 2506.01526 v1 pith:DU3XBDEZ submitted 2025-06-02 math.AG

Houari Benammar Ammar , Xi Chen , Nathan Grieve This is my paper

classification math.AG MSC 14J2914D06

keywords canonicallyfiberedsurfacesXiao'sconjecturerelativegenus5trigonalfibrationsgeometricboundNoether-typeinequalitiespseudo-canonicalmodelcubiccones

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper repairs gaps in an earlier claimed proof of Xiao's conjecture for canonically fibered surfaces whose general fibers have genus 5, simplifies the argument, and improves the final bound. The main theorem says that if such a surface has non-hyperelliptic general fibers, then the base curve $C$ is rational and the geometric genus satisfies $p_g(X)\le 50$. Combined with Sun's bound $p_g(X)\le 52$ for the hyperelliptic case, this settles Xiao's conjecture for relative genus 5. The same method also produces Noether-type lower bounds for $K_X^2$ of canonically fibered surfaces of relative genus 3 and 4.

What carries the argument

The central object is the pseudo relative canonical model $Y=\tau(X)$ for the map $\tau:X\dashrightarrow W=\mathbb{P}(f_*O_X(8\Gamma))^\vee$, where $\Gamma$ is the section appearing in the decomposition $K_X=f^*D+8\Gamma+V$. The proof classifies each fiber $Y_p$ into types A1, A2, or A3, and the key mechanism is the reducedness statement for type A3 fibers, which is proved separately in the nontrigonal and trigonal cases. In the trigonal case the paper constructs a resolution $R\to M$ of the family of cubic surfaces containing $Y$ and applies local lemmas about surfaces in $\Delta^3_{xyt}$ to rule out a fiber of the form $2J$ supported on a rational normal curve. The subsequent rank-2 statement for $f_*O_X(a\Gamma)$ and the regularity of the induced map to a ruled surface convert the geometric classification into the bound $\Gamma^2\le \frac1a(-d+\frac12\deg K_C)$, which yields the main theorem.

What would settle it

Check whether [Che17, Lemma 4.10] is proved independently in the present paper; if the lemma is not reproved and its conclusion fails for some canonically fibered surface, the proof that A3 fibers are reduced breaks down. Concretely, search for a genus-5 non-hyperelliptic fibration whose pseudo-canonical model has a cubic-cone fiber with multiplicity two along a rational normal curve; finding such a fiber would disprove Theorem 1.2.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1.2: for a canonically fibered surface $f:X\to C$ satisfying conditions C1--C6, if the general fibers have genus 5 and are non-hyperelliptic, then $g(C)=0$ and $p_g(X)\le 50$. The proof works through the pseudo-canonical model $Y$, the image of the rational map defined by $O_X(8\Gamma)$, and shows that each fiber $Y_p$ is an integral degree-$8$ genus-$5$ curve of one of three types: a complete intersection of three quadrics, a curve on a rational normal scroll, or a curve on a cubic cone. The hardest step is proving that $Y_p$ is reduced, especially for trigonal fibrations, and this is achieved through local analytic normal-form analyses near the vertex of the cubic cone. Once the model is understood, the paper proves that $f_*O_X(a\Gamma)$ has rank 2 for $a=4$ or $5$ and gives a regular map to a ruled surface, leading to the numerical bound on $p_g(X)$.

Load-bearing premise

The argument that a nonreduced A3 fiber cannot occur depends on [Che17, Lemma 4.10], a quoted local normal form at the vertex of the cubic cone, and that lemma comes from the same earlier paper whose proof the present paper says has gaps; if that lemma lies inside the gap, the exclusion of this fiber type would collapse.

Editorial extensions

If this is right

Xiao's conjecture for relative genus 5 is settled: any canonically fibered surface with general fiber genus 5 satisfies $p_g(X)\le 52$, with $p_g(X)\le 50$ in the non-hyperelliptic case.
Such a non-hyperelliptic surface must have rational base curve, $g(C)=0$.
For canonically fibered surfaces of relative genus 3, the bound $K_X^2\ge \frac{16}{3}d+\frac{10}{3}\deg K_C$ holds, with an analogous bound $K_X^2\ge \frac{22}{3}d+\frac{14}{3}\deg K_C$ for relative genus 4 under the stated hypotheses.
The paper replaces the erroneous structure theorems for the pseudo-canonical model in the earlier proof with a corrected trichotomy A1/A2/A3, thereby removing the incorrect assumption that the general fiber is a complete intersection of three quadrics.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The argument's dependence on the quoted local normal form [Che17, Lemma 4.10] suggests that a fully self-contained proof would need to reprove that lemma inside the new framework; the authors leave this verification implicit.
The same fiber-classification strategy might extend to relative genus 6, where Beauville's bound no longer applies, although a new fiber type beyond A1/A2/A3 would likely be required.
The bound $p_g(X)\le 50$ probably is not sharp, since the slack enters through the crude slope estimate in (2.15); a finer analysis of the section $\Gamma$ could lower the constant.
For genera 3 and 4, the inequalities point toward the asymptotic value of $K_X^2/p_g$ asked for in Xiao's problem list, and the proof identifies the extremal configurations (4.9) and (4.23) as the only cases needing special treatment.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The real news here is the trigonal case. The second author's earlier proof of Xiao's conjecture had a known gap: it assumed the canonical embedding is cut out by three quadrics, which fails for trigonal fibers. This paper works that case properly, and the payoff is concrete. Theorem 1.2 gives g(C)=0 and pg<=50 for non-hyperelliptic genus 5, which combined with Sun's hyperelliptic bound settles Xiao's Problem 6 with a global pg<=52. Theorem 1.3 adds Noether-type inequalities for relative genus 3 and 4. That is a genuine advance, not a repackaging.

The organization is transparent. The reduction of Theorem 1.2 to Theorem 2.1 and Corollary 2.2 is clear, the numerical chain from (2.3) to (2.15) checks out, and the paper is upfront that it is fixing gaps in [Che17]. The trigonal degeneration argument in Section 3.3 is long but self-contained, as far as I can tell, and the analytic lemmas 3.7–3.9 at least have plausible proofs.

Now the soft spot, and it is load-bearing. Section 3.2, the reducedness of Yp in the non-trigonal case, uses [Che17, Lemma 4.10] to assert a normal form at the vertex of the cubic cone. That lemma is not proved here. The paper claims independence from [Che17], but independence is exactly what is not established: if Lemma 4.10 lies inside the gap being fixed, the exclusion of a nonreduced A3 fiber collapses. The contradiction for m even and h(0,0)=0 really does use the derivative condition ∂g/∂x=0 from that lemma. This is not a minor gap. The trigonal case may stand on its own, but Theorem 1.2 as a whole depends on closing this import.

The analytic lemmas 3.4–3.6 I could not fully verify in a first reading; they are intricate, but they are at least in the paper. The external lemma is the one that bothers me.

Who is this for? Specialists in surface classification and relative canonical maps. It deserves a serious referee, but the referee needs to check Lemma 4.10 carefully, or the authors need to supply a proof in a revision. I would not cite Theorem 1.2 in my own work until that lemma is independently verified, but I would bring this paper to the reading group and I hope the editor sends it out.

Referee Report

2 major / 4 minor

Summary. The paper aims to complete the proof of Xiao's conjecture for canonically fibered surfaces of relative genus 5 by fixing gaps in the second author's earlier proof. The main theorem, Theorem 1.2, asserts that if f:X→C satisfies C1–C6 and the general fibers are non-hyperelliptic, then g(C)=0 and p_g(X)≤50; together with Sun's hyperelliptic bound p_g≤52 this would settle Xiao's conjecture. The proof reduces Theorem 1.2 to a structure theorem (Theorem 2.1) for the pseudo-canonical model Y of X/C, and then to Corollary 2.2, which gives a regular map from X to a ruled surface. The paper also proves Noether-type inequalities for canonically fibered surfaces of relative genus 3 and 4. The reduction of Theorem 1.2 to Theorem 2.1 and the numerical chain from (2.3) to (2.15) are clear and essentially self-contained. However, the proof of Theorem 2.1 in the non-trigonal case imports a key local normal-form lemma from [Che17] without proof, and this dependency is load-bearing.

Significance. If the proof is complete, the paper settles an open conjecture of Xiao for relative genus 5 and gives an explicit bound p_g≤52; this is a substantial result. The simplified proof of the non-trigonal case and the better bound p_g≤50 are genuine improvements over the earlier argument. The Noether-type inequalities for genus 3 and 4 are a useful addition. The paper also contains several interesting auxiliary results, including a general lemma on the arithmetic genus of proper transforms and a classification of linearly non-degenerate cubic surfaces in P^4. The main weakness is logical completeness: the proof of the crucial Theorem 2.1 depends on [Che17, Lemma 4.10], which is neither proved nor shown to lie outside the gaps that the paper claims to fix.

major comments (2)

[§3.2, paragraph beginning 'By [Che17, Lemma 4.10, p. 1048]'] The exclusion of a non-reduced A3 fiber in the non-trigonal case depends entirely on the imported assertion that, at the vertex o of the cubic cone, Y is locally isomorphic to {y^2=g(x,t)} with g(0,t)=t^δ and ∂g/∂x|_{x=0}≡0. This lemma is not proved in the present paper. The final contradiction for the case m even and h(0,0)=0 uses precisely the derivative condition ∂g/∂x|_{x=0}=0; if that condition is not available, the contradiction with smoothness of X_p at q disappears. Since Theorem 2.1 is the technical core of the proof of Theorem 1.2, and this step rules out the only nonreduced possibility in the non-trigonal case, [Che17, Lemma 4.10] is load-bearing. The authors should either provide a complete proof of this lemma or explicitly demonstrate, with details, that it lies outside the gaps in [Che17] that the paper claims to fix.
[§1, second paragraph] The statement that 'this paper is independent of [Che17]' is contradicted by the actual use of [Che17, Lemma 4.10] in §3.2. This is not merely a citation issue: the main theorem is not logically self-contained unless that lemma is proved or its correctness is fully established. The independence claim should be removed or qualified.

minor comments (4)

[§2.4, type A3 paragraph] The sentence 'So J must be singular at o. It follows that q≠o for q=J∩Λ' is not immediate as written; it relies on the previously established smoothness of Y_p at the point q=Y_p∩Λ, which in turn uses the normality of Y. This inference should be made explicit.
[§3.1, paragraph before (3.16)] In the description of generators of Pic(F3), the condition 'F^2=1' appears; for a Hirzebruch surface F3 one should have F^2=0, as stated correctly in Section 2.2. The same typo appears in the paragraph after (3.15) for the rational normal scroll case.
[§4.3, (4.35)–(4.37)] The use of Harris' theorem on theta characteristics to conclude h^0(O_{Y_p}(3q))=1 for every p from the value for general p should be stated more explicitly, specifying the parity result being invoked.
[§3.3, Lemmas 3.7 and 3.8] The strings 'tam', 'tam+1', and 't2m' appear to denote powers such as t^{am}, t^{am+1}, and t^{2m}; the typesetting should be corrected to avoid ambiguity.

Circularity Check

1 steps flagged · score 4.0 of 10

Non-trigonal A3 reducedness rests on an unproved self-cited normal-form lemma; the 'independent of [Che17]' claim is not supported.

self citation load bearing [Section 3.2, after Lemma 3.4, in the paragraph beginning 'By [Che17, Lemma 4.10, p. 1048]']
"By [Che17, Lemma 4.10, p. 1048], Y is locally isomorphic to the surface {y^2 = g(x, t)} ⊂ ∆^3_{xyt} for some g(x, t) ∈ C[[x, t]] satisfying (3.23) g(0, t) = t^δ and ∂g/∂x|_{x=0} ≡ 0, at the vertex o of the cone S for some δ ∈ Z+."

The proof of Theorem 2.1 in the non-trigonal A3 case reduces the exclusion of a non-reduced fiber Yp = 2J to the local structure of Y at the vertex o of the cubic cone S. The decisive contradiction for m even and h(0,0)=0 uses ∂g/∂x|_{x=0}=0, which is exactly the content of the imported [Che17, Lemma 4.10]. That lemma is not proved in the present paper, and the paper explicitly says it is fixing gaps in [Che17] and claims to be 'independent of [Che17]'. No argument is given that Lemma 4.10 lies outside the gaps being fixed. Thus a load-bearing premise rests on an unverified self-citation rather than on an independent derivation; if the lemma fails or lies inside the gap, the non-reduced A3 fiber is not ruled out. This is not a fitted-input circularity, but it is load-bearing self-citation.

full rationale

The central claim, Theorem 1.2, is an external benchmark (a bound on pg(X) for canonically fibered surfaces of relative genus 5), and the main derivation chain does not assume Xiao's conjecture or the desired bound. No fitted constants are renamed as predictions. However, the paper is not fully self-contained: Section 3.2 imports [Che17, Lemma 4.10] to establish the local normal form at the vertex of the cubic cone, and the contradiction excluding a non-reduced A3 fiber in the non-trigonal case depends on the derivative condition in that lemma. The paper states that it is fixing gaps in [Che17] and also claims independence from [Che17], but it does not prove Lemma 4.10 or show that the lemma lies outside the gap being fixed. This is a load-bearing self-citation from the very paper whose proof is being repaired. Since the rest of the argument, including the derivation of Corollary 2.2 from Theorem 2.1 and Theorem 1.2 from Corollary 2.2, is internally carried out, the appropriate score is 4 rather than a higher value.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central theorem is a pure existence and bound statement; no constants are fitted to data. The main nontrivial imports are standard theorems plus one lemma from [Che17] that is not reproved here. The pseudo-canonical model Y is a construction from [Che17], not a genuinely new postulated entity.

assumptions (5)

standard math Logarithmic Bogomolov-Miyaoka-Yau inequality applied to (X, Gamma).
Used in Section 2.1 to derive the lower bound (2.3) on Gamma^2, a key numerical input.
standard math Xiao's structure results for canonically fibered surfaces: C is rational or elliptic, N=8Gamma+V, and f_* omega_X = L direct sum M with M tensor omega_C^{-1} semipositive.
Quoted from [Xia85] in Section 2.1 and used throughout the proof.
standard math Saint-Donat's classification of canonical genus 5 curves, including the trigonal case lying on rational normal scrolls.
Used in Section 2.2 to assign types A1, A2, A3 and in the trigonal case.
ad hoc to paper [Che17, Lemma 4.10] local normal form y^2=g(x,t) with g(0,t)=t^delta and partial g/partial x=0 at x=0 at the vertex of the cubic cone.
Invoked in Section 3.2 to rule out the nonreduced A3 fiber; not proved in this paper and sourced from the flawed paper whose gaps this work claims to fix.
standard math Harris's parity theorem for theta characteristics on canonical curves.
Used in Corollary 2.2 and Section 4.3 to force h^0=1 from parity and dimension bounds.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Trigonal Canonically Fibered Surfaces." pith.science (2026). https://pith.science/paper/DU3XBDEZ

@misc{pith2026250601526,
  author       = {Pith},
  title        = {Pith review of: Trigonal Canonically Fibered Surfaces},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DU3XBDEZ}},
  note         = {Machine review of arXiv:2506.01526}
}

read the original abstract

We fix some gaps of a proof of Xiao's conjecture on canonically fibered surfaces of relative genus 5 by the second author. Our argument simplifies the original proof and gives a much better bound on the geometric genus of the surface. Also we apply the same argument to canonically fibered surfaces of relative genus 3 and 4 to obtain some Noether-type inequalities for these surfaces.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

10 extracted references · 10 canonical work pages

[1]

L'application canonique pour les surfaces de type g\' e n\' e ral

Arnaud Beauville. L'application canonique pour les surfaces de type g\' e n\' e ral. Invent. Math. , 55(2):121--140, 1979

work page 1979
[2]

Xiao's conjecture on canonically fibered surfaces

Xi Chen. Xiao's conjecture on canonically fibered surfaces. Advances in Mathematics , 322:1033--1084, 2017

work page 2017
[3]

Lectures on Vanishing Theorems

H\'el\`ene Esnault and Eckart Viehweg. Lectures on Vanishing Theorems . DMV Seminar Vol 20. Springer-Verlag, Berlin-New York, 1992

work page 1992
[4]

Theta-characteristics on algebraic curves

Joe Harris. Theta-characteristics on algebraic curves. Transactions of the American Mathematical Society , 271(2):611--638, 1982

work page 1982
[5]

Canonically fibered surfaces of general type

Xin Lu. Canonically fibered surfaces of general type. J. Inst. Math. Jussieu , 19(1):209--229, 2020

work page 2020
[6]

The minimal number of quotient singularities on surfaces with given numerical invariants

Yoichi Miyaoka. The minimal number of quotient singularities on surfaces with given numerical invariants. Math. Ann. , 268:159--171, 1984

work page 1984
[7]

Saint-Donat

B. Saint-Donat. On petri's analysis of the linear system of quadrics through a canonical curve. Mathematische Annalen , 206(2):157--175, Jun 1973

work page 1973
[8]

On canonical fibrations of algebraic surfaces

Xiaotao Sun. On canonical fibrations of algebraic surfaces. manuscripta mathematica , 83(1):161--169, Dec 1994

work page 1994

Show all 10 references

[9]

G. Xiao. L'irrégularité des surfaces de type général dont le système canonique est composé d'un pinceau. Compositio Mathematica , 56(2):251--257, 1985

1985
[10]

G. Xiao. Problem list in birational geometry of algebraic varieties: open problems. in: XXIII International Symposium, Division of Mathematics, the Taniguchi Foundation. , pages pp. 36--40, 1988

1988

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.

[1] [1]

L'application canonique pour les surfaces de type g\' e n\' e ral

Arnaud Beauville. L'application canonique pour les surfaces de type g\' e n\' e ral. Invent. Math. , 55(2):121--140, 1979

work page 1979

[2] [2]

Xiao's conjecture on canonically fibered surfaces

Xi Chen. Xiao's conjecture on canonically fibered surfaces. Advances in Mathematics , 322:1033--1084, 2017

work page 2017

[3] [3]

Lectures on Vanishing Theorems

H\'el\`ene Esnault and Eckart Viehweg. Lectures on Vanishing Theorems . DMV Seminar Vol 20. Springer-Verlag, Berlin-New York, 1992

work page 1992

[4] [4]

Theta-characteristics on algebraic curves

Joe Harris. Theta-characteristics on algebraic curves. Transactions of the American Mathematical Society , 271(2):611--638, 1982

work page 1982

[5] [5]

Canonically fibered surfaces of general type

Xin Lu. Canonically fibered surfaces of general type. J. Inst. Math. Jussieu , 19(1):209--229, 2020

work page 2020

[6] [6]

The minimal number of quotient singularities on surfaces with given numerical invariants

Yoichi Miyaoka. The minimal number of quotient singularities on surfaces with given numerical invariants. Math. Ann. , 268:159--171, 1984

work page 1984

[7] [7]

Saint-Donat

B. Saint-Donat. On petri's analysis of the linear system of quadrics through a canonical curve. Mathematische Annalen , 206(2):157--175, Jun 1973

work page 1973

[8] [8]

On canonical fibrations of algebraic surfaces

Xiaotao Sun. On canonical fibrations of algebraic surfaces. manuscripta mathematica , 83(1):161--169, Dec 1994

work page 1994