Closed Immersion¶
A scheme morphism that identifies its source with a closed subscheme through locally surjective maps from ambient functions.
Core Idea¶
A closed immersion \(i:Z\to X\) is a morphism of schemes that realizes \(Z\) as a closed subscheme of \(X\). Topologically, it identifies \(Z\) with a closed subset of \(X\); structurally, the map \(\mathcal O_X\to i_*\mathcal O_Z\) is surjective. Equivalently, on each affine open \(\operatorname{Spec}R\subset X\), the inverse image is \(\operatorname{Spec}(R/I)\) for an ideal \(I\subset R\). Thus ambient functions restrict to source functions and local equations—encoded by an ideal sheaf—determine more than the set of points.[1][2]
The Stacks Project's initial definition includes a locally generated kernel; for morphisms of schemes its Lemma 29.2.1 proves equivalent formulations using closed image plus sheaf surjectivity and local quotient rings. These are alternate tests for one identity, not independent extra requirements that must be piled onto the theorem.[1]
Structural Signature¶
Sig role-phrases: source-and-ambient schemes → closed topological placement → surjective structure-sheaf map → local quotient/ideal description.
- Source and ambient schemes: \(Z\) and \(X\) supply underlying spaces, structure sheaves and the morphism being classified. A continuous map alone lacks this information.[1]
- Closed topological placement: \(i\) is a homeomorphism onto a closed subset, locating the source points in the target.[1]
- Quotient structure: \(\mathcal O_X\to i_*\mathcal O_Z\) is surjective, locally equivalent to \(R\to R/I\). A closed image without this condition is not enough.[1]
- Defining ideal: the kernel records which ambient functions vanish in the source. The ideal sheaf is an equivalent representation of the closed subscheme and retains nonreduced thickness invisible to the support.[1][2]
Base-change and composition stability are consequences proved for the class, not extra constitutive roles.[1]
What It Is Not¶
It is not merely an injective continuous map with closed image. Two morphisms can have the same image of points but different quotient structure sheaves. Nor is it restricted to reduced varieties: \(\operatorname{Spec}(k[t]/(t^2))\to\operatorname{Spec}(k[t])\) is a closed immersion defined by \((t^2)\), while \((t)\) defines the reduced origin with the same underlying point.[1][2]
It is also not every scheme morphism called an “embedding” informally. The sheaf/local-ring quotient criterion must be checked; a visual inclusion is insufficient.[1]
Scope of Application¶
Affine algebra gives the basic case \(\operatorname{Spec}(R/I)\to\operatorname{Spec}R\). The ideal \(I\) can define a reduced plane curve, a nonreduced thickened point or another closed subscheme. Stacks Lemma 26.8.2 identifies affine closed immersions with ideals, and Lemma 29.2.1 extends the quotient criterion across affine charts of a general target.[2][1]
Projective geometry supplies a different realization. The Segre map \(\mathbf P^n\times\mathbf P^m\to\mathbf P^{nm+n+m}\) is proved to be a closed immersion by checking that the relevant affine-chart coordinate maps are surjective. Its source is a product of projective schemes rather than an affine thickened point, but the local quotient structure remains.[3]
Clarity¶
Distinguish the closed support from the closed subscheme structure. The ideals \((t)\) and \((t^2)\) in \(k[t]\) give quotient schemes supported at the same origin point. The second quotient retains a nonzero nilpotent class of \(t\); a point-set picture cannot recover it. A claim that “the image is closed” therefore answers only the topological half of the classification.[2]
Also distinguish an ideal from its generating equation list: different generators may describe one ideal, while distinct ideals can share a vanishing set. The closed-immersion identity follows the quotient/ideal sheaf, not the display of a particular polynomial.[1]
Manages Complexity¶
The local quotient criterion reduces a sheaf-theoretic question to familiar ring maps on affine charts. Instead of manipulating the entire embedded space at once, inspect ideals \(I\subset R\) and verify compatible local descriptions. Stacks proves that these descriptions characterize the global morphism and that a quasi-coherent ideal sheaf determines a closed subscheme.[1][2]
The reduction preserves essential complexity: nilpotents and multiplicities remain visible in \(R/I\), even when they disappear from the underlying set. Composition and base change retain the closed-immersion class, making the construction stable under common geometric operations.[1]
Abstract Reasoning¶
Given a proposed \(i:Z\to X\), choose any affine open \(U=\operatorname{Spec}R\) of \(X\). If \(i^{-1}(U)\) is \(\operatorname{Spec}(R/I)\) over \(U\) for an ideal \(I\), and this holds on all such charts, Stacks Lemma 29.2.1 licenses the conclusion that \(i\) is a closed immersion. Conversely, a known closed immersion lets one recover these quotient presentations and the kernel ideal sheaf.[1]
The ideal also supports relational reasoning: Stacks Lemma 29.2.2 says factorization of one closed immersion through another corresponds to reverse inclusion of their defining ideals. Thus \((t^2)\subset(t)\) reflects a thicker-to-thinner quotient relationship even though the supports agree.[1]
Knowledge Transfer¶
The affine thickened origin and the projective Segre map share scheme morphism / closed placement / locally surjective coordinate map / defining ideal. In the affine example one global quotient \(k[t]\to k[t]/(t^2)\) displays the structure. In the projective example different affine charts carry the local surjections; gluing recovers the global closed embedding.[2][3]
The transferable lesson is not “every injective geometric picture is a closed immersion.” One must transport the ringed-space structure, including nilpotents where present. The same-looking set image need not transport the same subscheme.[1]
Examples¶
Nonreduced affine thickened origin. Mapped back: ambient = \(\operatorname{Spec}k[t]\); source = \(\operatorname{Spec}(k[t]/(t^2))\); closed image = the origin; quotient map = \(k[t]\twoheadrightarrow k[t]/(t^2)\); ideal = \((t^2)\). It has the same underlying support as \(\operatorname{Spec}(k[t]/(t))\) but different scheme structure. This is a direct instance of the Stacks affine-quotient lemma.[2]
Segre product embedding. Mapped back: ambient = \(\mathbf P^{nm+n+m}\); source = \(\mathbf P^n\times\mathbf P^m\); closed image = the Segre image; quotient test = surjective coordinate-ring maps on an affine cover; defining ideal = the kernel of those maps, without requiring one global affine chart. Stacks proves this is a closed immersion.[3]
Structural Tensions¶
Visible support versus hidden thickness. Using only the point-set image makes geometry easy to visualize but erases nilpotents; retaining ideals distinguishes reduced and thickened embeddings but requires sheaf/ring data. Neither a picture alone nor a bare equation list gives the whole classification. Diagnostic: Which ideal defines the source, not merely which points does it occupy?[2]
Local quotient convenience versus global compatibility. One affine chart makes surjectivity concrete, but an arbitrary target requires compatible quotients across a cover. Checking only one chart risks a false global conclusion; carrying the ideal sheaf costs bookkeeping but secures a coherent closed subscheme. Diagnostic: Do the local quotient descriptions agree on overlaps?[1]
Structural–Framed Character¶
Evaluative weight. Closed immersion is a formal mathematical classification, not a judgment of geometric desirability. Human-practice dependence. Mathematicians choose schemes, coordinates and proofs, yet once the morphism is fixed, closed image and sheaf surjectivity are rule-governed properties, not discretionary labels.[1]
Institutional origin. The term belongs to algebraic-geometry practice and is standardized in references such as the Stacks Project; no single institution's notation is constitutive. Vocabulary travel. “Embedding” travels broadly, but the closed immersion criterion transfers literally only where scheme structure, sheaves and local quotients are preserved.[1]
Import versus recognition. A new morphism can be recognized by the affine quotient test regardless of the equations used to introduce it. Calling a closed point-set inclusion in another setting a “closed immersion” imports a metaphor unless its ringed-space conditions are present. Its character: strongly structural within scheme theory, with a domain-bound mathematical carrier.[1]
Structural Core vs. Domain Accent¶
Portable skeleton. A generic idea of faithful inclusion into a larger object may be a future-prime question, but this wave does not claim or add that prime. The immediate necessary live genus is Morphism of schemes: Stacks explicitly starts with a morphism of schemes and adds closed-image/quotient conditions. The proposed edge is strict subsumption under that domain-specific genus, not under a merely topical prime.[1]
Domain-bound mechanism. The source is identified with a closed subset while its structure sheaf is a local quotient of the ambient sheaf. Ideals record equations and nilpotent thickness; affine charts certify the condition, and projective charts glue it. This precise ringed-space mechanism is what the examples share.[1][2][3]
Why not prime. Without schemes, structure sheaves, affine coordinate rings and ideals, the defining equivalence cannot even be stated. A faithful inclusion of records or a closed subset of a topological space may resemble the shape but is not this identity. The broad inclusion skeleton could travel, while Closed Immersion remains domain-specific to algebraic geometry.
Instantiates / Related Primes¶
This entry is a kind of Morphism of schemes.
The staged typed relation is strict subsumption under Morphism of schemes. No canonical DAG edge has been applied. Fiber Product of Schemes is related to base change, under which closed immersions are stable, but it is not a necessary genus. A generic prime “embedding” was not used as a parent because it would not establish the quotient-sheaf condition.[1]
Relationships to Other Abstractions¶
Current abstraction Closed Immersion Domain-specific
Parents (1) — more general patterns this builds on
-
Closed Immersion is a kind of Morphism of schemes Domain-specific
A closed immersion is a morphism of schemes satisfying additional closed-image and quotient-sheaf conditions.The Stacks Project defines it as a scheme morphism and proves equivalence with local affine quotient maps.
Hierarchy path (1) — routes to 1 parentless root
- Closed Immersion → Morphism of schemes → Relation
Neighborhood in Abstraction Space¶
Closed Immersion sits in a moderately populated region (53rd percentile for distinctiveness): it has near-neighbors but no dense thicket of look-alikes.
Family — Categories, Sheaves & Homotopy (21 abstractions)
Nearest neighbors
- Ringed Space — 0.88
- Locally Closed Subset — 0.86
- Dévissage — 0.85
- Sheaf — 0.85
- Quasi-projective variety — 0.85
Computed from structural-signature embeddings · 2026-10-08
Not to Be Confused With¶
Closed topological embedding need not include source structure sheaf as an ambient quotient. Open immersion identifies a different local type of subscheme. Reduced closed subset may have the same support as a thicker nonreduced closed subscheme. Finite morphism is a broader property that does not by itself establish the closed-immersion identity.[1]
References¶
[1] The Stacks Project, §29.2 “Closed immersions,” especially Lemma 29.2.1 (equivalent definitions), Lemma 29.2.2 (ideal/factorization), and Lemmas 29.2.4–29.2.5 (base change and composition). registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j ↩k ↩l ↩m ↩n ↩o ↩p ↩q ↩r ↩s ↩t ↩u ↩v ↩w ↩x
[2] The Stacks Project, Lemma 26.8.2 “Closed subspaces of affine schemes,” identifying affine closed immersions with ideals and quotient rings. registry ↩a ↩b ↩c ↩d ↩e ↩f ↩g ↩h ↩i ↩j
[3] The Stacks Project, Lemma 27.13.6 “Segre embedding,” including its affine-chart proof of closed immersion. registry ↩a ↩b ↩c ↩d