Relative Effective Cartier Divisors

• Relative effective Cartier divisors are a family of divisor-theoretic constructs defined via invertible submodules over ring extensions and localized cohomology.
• They extend classical Cartier divisors by integrating étale, Zariski, and Picard group techniques to measure descent and connectedness properties.
• Their applications span deformation theory, arithmetic effectivity, and singular surface analysis, providing both abstract insights and algorithmic methods.

Searching arXiv for recent and foundational papers on relative Cartier divisors, effective Cartier divisors, and related divisor-theoretic frameworks. Relative effective Cartier divisors appear in several technically distinct frameworks. For a commutative ring extension $A\subset B$, the established relative object is the group $I(A,B)$ of invertible $A$-submodules of $B$, which the literature identifies as the group of relative Cartier divisors; when $A$ is a domain with fraction field $K$, $I(A,K)$ recovers the usual Cartier divisor group (Sadhu et al., 2015). In deformation theory, one studies an effective divisor $D\subset X$ and asks whether $D$, or some multiple $nD$, lifts as an effective Cartier divisor along a first-order deformation of $I(A,B)$0 (Biswas et al., 2020). In arithmetic effectivity, Cartier divisors on a geometrically integral projective variety over a number field are encoded by presentations that make local Weil functions effectively computable, even when the divisor is defined only after passage to a finite extension (Grieve, 21 Oct 2025). Taken together, these viewpoints show that “relative effective Cartier divisors” do not denote a single universal construction, but rather a family of divisor-theoretic formalisms organized around invertible submodules, pullback data, deformation along a base, and effective computation.

1. Ring extensions and the basic relative divisor group

For a commutative ring extension $I(A,B)$1, the paper "Relative Cartier Divisors and Laurent Polynomial Extensions" defines

$I(A,B)$2

under multiplication, and states that an invertible $I(A,B)$3-submodule is also said to be a relative Cartier divisor (Sadhu et al., 2015). The definition is multiplicative: the $I(A,B)$4-submodules of $I(A,B)$5 form a monoid under multiplication with identity $I(A,B)$6, and $I(A,B)$7 is invertible if $I(A,B)$8 for some $I(A,B)$9. This is the basic ring-theoretic relative divisor group.

The same source gives a sheaf-theoretic interpretation. Writing $A$0, the group $A$1 is described by a Zariski-cohomological formula of the form

$A$2

stated more precisely there using the quotient sheaf of units (Sadhu et al., 2015). This realizes relative Cartier divisors as global sections of a quotient measuring the difference between the structure sheaf of the base and its direct image from the extension.

A fundamental structural sequence relates units, relative Cartier divisors, and Picard groups: $A$3 This exact sequence is the basic interface between divisor theory and line bundles in the relative setting (Sadhu et al., 2015). A plausible implication is that $A$4 should be viewed as the divisor-theoretic correction term that measures how far units and line bundles fail to descend across the extension.

The same ring-extension formalism already contains the ordinary Cartier divisor group as a special case. When $A$5 is a domain with fraction field $A$6, the identification $A$7 with the usual Cartier divisor group gives a direct relative-to-absolute bridge (Sadhu et al., 2015). This point is conceptually important: the relative theory is not merely analogous to ordinary divisor theory, but extends it.

2. Bass contraction and Laurent polynomial extensions

A central structural result is that the functor $A$8 behaves as a Bass contracted functor on ring extensions. For a functor $A$9 on ring extensions, the contraction $B$0 is defined as the cokernel of

$B$1

and one has the associated sequence

$B$2

A functor is contracted if this sequence is naturally split exact (Sadhu et al., 2015).

For the relative Cartier divisor functor $B$3, the main theorem states that $B$4 is contracted and that

$B$5

In particular,

$B$6

and the invariance

$B$7

also holds (Sadhu et al., 2015). The same paper proves the vanishing statements

$B$8

This decomposition isolates the genuinely new contribution created by Laurent polynomial extension. The paper interprets the quotient sheaf $B$9 as measuring the defect of the map on connected components or idempotent structure, and identifies its global étale sections with $A$0 (Sadhu et al., 2015). The result is therefore not only formal: it exhibits relative Cartier divisors as controlled by an étale-cohomological invariant.

The étale topology is essential in this description. The same source explicitly notes that the corresponding Zariski analogue fails in general, whereas one does have

$A$1

On a henselian local base,

$A$2

which identifies the stalkwise content of the sheaf-theoretic description (Sadhu et al., 2015). This gives the relative divisor group a cohomological structure that is sensitive to connectedness rather than merely local principal generation.

Several consequences sharpen the structure theory. The paper states that $A$3 is always torsion-free; if $A$4 is pseudo-geometric and finite-dimensional, then $A$5 is free abelian; if $A$6 is connected, then $A$7 iff $A$8 is connected and $A$9 is injective; and if $K$0 is finite and connected, then $K$1 (Sadhu et al., 2015). It also relates vanishing to seminormality and anodality: $K$2 is equivalent to $K$3 being seminormal in $K$4, and if $K$5, then $K$6 is anodal (Sadhu et al., 2015).

3. Scheme-theoretic relative Picard groups and relative $K$7-theory

A scheme-theoretic formulation is given by the relative Picard group $K$8 of a morphism $K$9. Bass defines $I(A,K)$0 as the abelian group generated by symbols

$I(A,K)$1

where $I(A,K)$2 are line bundles on $I(A,K)$3 and

$I(A,K)$4

is an isomorphism, subject to additivity, composition, and triviality when $I(A,K)$5 comes from the base (Sadhu et al., 2016). There is a natural exact sequence

$I(A,K)$6

The key identification occurs for faithful affine morphisms, meaning affine $I(A,K)$7 such that

$I(A,K)$8

is injective (Sadhu et al., 2016). In the affine case $I(A,K)$9, this is exactly the condition $D\subset X$0. For such $D\subset X$1, the paper recalls that

$D\subset X$2

and deduces

$D\subset X$3

Lemma 1.2 there gives an explicit isomorphism

$D\subset X$4

sending an invertible submodule $D\subset X$5 to the class $D\subset X$6, where $D\subset X$7 (Sadhu et al., 2016). In this form, a relative Cartier divisor is literally a line subbundle of $D\subset X$8 compatible with the $D\subset X$9-structure.

The same paper embeds relative Cartier divisors into relative $D$0-theory. For a ring map $D$1, the relative Grothendieck group $D$2 is defined via triples $D$3, and one has the exact sequence

$D$4

The $D$5-operations make $D$6 into a special $D$7-ring with positive structure, and the top two stages of the $D$8-filtration satisfy

$D$9

with

$nD$0

Equivalently, relative Cartier divisors are the top quotient for the $nD$1-filtration on relative $nD$2 (Sadhu et al., 2016).

This identification becomes sharper in special cases. For subintegral extensions $nD$3, Proposition 2.5 states

$nD$4

together with an exact sequence

$nD$5

A plausible implication is that under subintegrality, the entire relative $nD$6 group is governed by relative Cartier divisors (Sadhu et al., 2016).

4. Effectivity in families: deformation and local obstruction

In deformation theory, the phrase “effective Cartier divisor” appears in a different relative sense. The paper "Local topological obstruction for divisors" studies a smooth projective family

$nD$7

with $nD$8 a complex manifold, base point $nD$9, central fiber $I(A,B)$00, and an effective divisor $I(A,B)$01. The problem is to determine, for a tangent direction $I(A,B)$02, when $I(A,B)$03 deforms along the first-order deformation $I(A,B)$04, when the class $I(A,B)$05 deforms as a Hodge class, and how the two obstructions compare (Biswas et al., 2020).

The paper refines the classical obstruction theory by replacing the global group $I(A,B)$06 with a local cohomology group supported on $I(A,B)$07. Writing $I(A,B)$08 and $I(A,B)$09, one has the exact sequence

$I(A,B)$10

and, for locally free $I(A,B)$11,

$I(A,B)$12

Choosing local equations $I(A,B)$13 for $I(A,B)$14, the forms

$I(A,B)$15

glue to a global section

$I(A,B)$16

whose image is the usual cohomology class $I(A,B)$17 (Biswas et al., 2020).

The paper defines contraction maps

$I(A,B)$18

inducing

$I(A,B)$19

$I(A,B)$20

and the natural map

$I(A,B)$21

For $I(A,B)$22,

$I(A,B)$23

Thus $I(A,B)$24 is a local refinement of the ordinary Hodge-theoretic obstruction (Biswas et al., 2020).

The main theorem states that for every tangent vector $I(A,B)$25,

$I(A,B)$26

as an effective Cartier divisor (Biswas et al., 2020). This is a genuinely relative effective Cartier divisor statement: vanishing of a local topological obstruction is equivalent to lifting some positive multiple of the divisor along the infinitesimal base direction. The same paper defines the geometric obstruction map

$I(A,B)$27

describes saturation by the criterion

$I(A,B)$28

and proves that there are first-order deformations $I(A,B)$29 for which $I(A,B)$30 deforms as a Hodge class but $I(A,B)$31 does not lift as an effective Cartier divisor (Biswas et al., 2020). This removes a common oversimplification: deformation of the cohomology class does not in general coincide with deformation of the divisor itself.

5. Effective calculation via presentations of Cartier divisors

A different meaning of effectivity arises in arithmetic geometry. The paper "Effective calculation of local Weil functions via presentations of Cartier divisors" works with a geometrically integral projective variety $I(A,B)$32, a Cartier divisor $I(A,B)$33 on $I(A,B)$34, and a finite extension $I(A,B)$35 over which $I(A,B)$36 is defined. It does not develop a new theory of Cartier divisors in a relative scheme-theoretic sense; rather, it uses presentations of $I(A,B)$37 to make local Weil functions effectively computable (Grieve, 21 Oct 2025).

The key preparatory lemma is that the line bundle of $I(A,B)$38 can be written as

$I(A,B)$39

with $I(A,B)$40 and $I(A,B)$41 globally generated over $I(A,B)$42. The proof is effective in spirit: take a very ample line bundle $I(A,B)$43, then for $I(A,B)$44 one may choose

$I(A,B)$45

for example with $I(A,B)$46, the Castelnuovo–Mumford regularity (Grieve, 21 Oct 2025). A presentation of $I(A,B)$47 is then data

$I(A,B)$48

where $I(A,B)$49 is a meromorphic section of $I(A,B)$50, and

$I(A,B)$51

are global generating sections of $I(A,B)$52 and $I(A,B)$53, respectively.

From such a presentation, the paper defines the local Weil function at a place $I(A,B)$54 by

$I(A,B)$55

for $I(A,B)$56 (Grieve, 21 Oct 2025). Special cases make the formalism explicit. If $I(A,B)$57 and $I(A,B)$58, then

$I(A,B)$59

with additivity

$I(A,B)$60

If $I(A,B)$61 have presentations $I(A,B)$62, then the induced presentation of $I(A,B)$63 satisfies

$I(A,B)$64

For a degree $I(A,B)$65 hypersurface in $I(A,B)$66 defined by a homogeneous form $I(A,B)$67, a natural presentation using $I(A,B)$68 and $I(A,B)$69 yields an explicit projective-space formula involving degree-$I(A,B)$70 monomials (Grieve, 21 Oct 2025).

The main theorem states that if $I(A,B)$71 and $I(A,B)$72 are two presentations of the same Cartier divisor $I(A,B)$73, then for each place $I(A,B)$74 there exists an effectively computable constant $I(A,B)$75 such that

$I(A,B)$76

The proof forms the difference presentation

$I(A,B)$77

reduces to bounding expressions

$I(A,B)$78

and then uses a projective embedding, affine charts $I(A,B)$79, bounded subsets

$I(A,B)$80

polynomial descriptions on refinements $I(A,B)$81, and an effective Hilbert Nullstellensatz producing relations

$I(A,B)$82

with effective bounds (Grieve, 21 Oct 2025). In this sense, the paper provides an algorithmic presentation of Cartier divisors defined over finite extensions, rather than a general theory of relative effective Cartier divisors.

6. Singular surfaces, almost Cartier divisors, and limits of effectivity

On singular surfaces, divisor theory requires further refinement. The paper "Divisors class groups of singular surfaces" works with generalized divisors as fractional ideal sheaves on schemes satisfying $I(A,B)$83 and $I(A,B)$84. A generalized divisor is a coherent $I(A,B)$85-submodule

$I(A,B)$86

that is nondegenerate and reflexive; it is principal if generated by a single nonzero-divisor, Cartier if locally principal everywhere, and almost Cartier if locally principal off a subset of codimension at least $I(A,B)$87 (Hartshorne et al., 2013). The corresponding class groups are

$I(A,B)$88

A divisor is effective if

$I(A,B)$89

and then determines a codimension-one subscheme $I(A,B)$90 without embedded components (Hartshorne et al., 2013).

For a normalization $I(A,B)$91 with conductor subscheme $I(A,B)$92 and inverse image $I(A,B)$93, the local description of almost Cartier divisors is

$I(A,B)$94

under the smoothness and conductor hypotheses stated there (Hartshorne et al., 2013). Globally, one has exact sequences

$I(A,B)$95

and, when $I(A,B)$96 is smooth,

$I(A,B)$97

These formulas show that on singular surfaces, Cartier-type data is controlled by normalization together with conductor geometry (Hartshorne et al., 2013).

A nearby but logically distinct issue is positivity. The paper "A question on effective strictly nef divisors" asks whether every effective strictly nef Cartier divisor on a projective variety is big and shows that the answer is negative in every dimension $I(A,B)$98 (Fontanari, 2023). The counterexample in the appendix constructs an effective prime divisor

$I(A,B)$99

that is strictly nef but not big, with

$A$00

(Fontanari, 2023). This is relevant because “effective Cartier divisor” does not by itself impose positivity properties such as bigness, even in projective geometry.

These adjacent theories delimit the scope of the subject. Relative Cartier divisors in the sense of invertible submodules, effective Cartier divisors in deformation problems, and effective computation for Cartier divisors over finite extensions are all robust frameworks, but they address different structural questions. A plausible implication is that any encyclopedia-level treatment must keep these meanings separate: relative divisor theory is cohomological and functorial, deformation-theoretic effectivity is controlled by local obstruction groups, and computational effectivity is achieved by explicit presentations rather than by enlarging the abstract definition of relative Cartier divisor.