Lemma 66.20.1. Let $S$ be a scheme. Let $f : X \to Y$ be a representable morphism of algebraic spaces over $S$. Then $f$ is affine (in the sense of Section 66.3) if and only if for all affine schemes $Z$ and morphisms $Z \to Y$ the scheme $X \times _ Y Z$ is affine.

## 66.20 Affine morphisms

We have already defined in Section 66.3 what it means for a representable morphism of algebraic spaces to be affine.

**Proof.**
This follows directly from the definition of an affine morphism of schemes (Morphisms, Definition 29.11.1).
$\square$

This clears the way for the following definition.

Definition 66.20.2. Let $S$ be a scheme. Let $f : X \to Y$ be a morphism of algebraic spaces over $S$. We say $f$ is *affine* if for every affine scheme $Z$ and morphism $Z \to Y$ the algebraic space $X \times _ Y Z$ is representable by an affine scheme.

Lemma 66.20.3. Let $S$ be a scheme. Let $f : X \to Y$ be a morphism of algebraic spaces over $S$. The following are equivalent:

$f$ is representable and affine,

$f$ is affine,

for every affine scheme $V$ and étale morphism $V \to Y$ the scheme $X \times _ Y V$ is affine,

there exists a scheme $V$ and a surjective étale morphism $V \to Y$ such that $V \times _ Y X \to V$ is affine, and

there exists a Zariski covering $Y = \bigcup Y_ i$ such that each of the morphisms $f^{-1}(Y_ i) \to Y_ i$ is affine.

**Proof.**
It is clear that (1) implies (2), that (2) implies (3), and that (3) implies (4) by taking $V$ to be a disjoint union of affines étale over $Y$, see Properties of Spaces, Lemma 65.6.1. Assume $V \to Y$ is as in (4). Then for every affine open $W$ of $V$ we see that $W \times _ Y X$ is an affine open of $V \times _ Y X$. Hence by Properties of Spaces, Lemma 65.13.1 we conclude that $V \times _ Y X$ is a scheme. Moreover the morphism $V \times _ Y X \to V$ is affine. This means we can apply Spaces, Lemma 64.11.5 because the class of affine morphisms satisfies all the required properties (see Morphisms, Lemmas 29.11.8 and Descent, Lemmas 35.23.18 and 35.37.1). The conclusion of applying this lemma is that $f$ is representable and affine, i.e., (1) holds.

The equivalence of (1) and (5) follows from the fact that being affine is Zariski local on the target (the reference above shows that being affine is in fact fpqc local on the target). $\square$

Lemma 66.20.4. The composition of affine morphisms is affine.

**Proof.**
Omitted. Hint: Transitivity of fibre products.
$\square$

Lemma 66.20.5. The base change of an affine morphism is affine.

**Proof.**
Omitted. Hint: Transitivity of fibre products.
$\square$

Lemma 66.20.6. A closed immersion is affine.

**Proof.**
Follows immediately from the corresponding statement for morphisms of schemes, see Morphisms, Lemma 29.11.9.
$\square$

Lemma 66.20.7. Let $S$ be a scheme. Let $X$ be an algebraic space over $S$. There is an anti-equivalence of categories

which associates to $f : Y \to X$ the sheaf $f_*\mathcal{O}_ Y$. Moreover, this equivalence is compatible with arbitrary base change.

**Proof.**
This lemma is the analogue of Morphisms, Lemma 29.11.5. Let $\mathcal{A}$ be a quasi-coherent sheaf of $\mathcal{O}_ X$-algebras. We will construct an affine morphism of algebraic spaces $\pi : Y = \underline{\mathop{\mathrm{Spec}}}_ X(\mathcal{A}) \to X$ with $\pi _*\mathcal{O}_ Y \cong \mathcal{A}$. To do this, choose a scheme $U$ and a surjective étale morphism $\varphi : U \to X$. As usual denote $R = U \times _ X U$ with projections $s, t : R \to U$. Denote $\psi : R \to X$ the composition $\psi = \varphi \circ s = \varphi \circ t$. By the aforementioned lemma there exists an affine morphisms of schemes $\pi _0 : V \to U$ and $\pi _1 : W \to R$ with $\pi _{0, *}\mathcal{O}_ V \cong \varphi ^*\mathcal{A}$ and $\pi _{1, *}\mathcal{O}_ W \cong \psi ^*\mathcal{A}$. Since the construction is compatible with base change there exist morphisms $s', t' : W \to V$ such that the diagrams

are cartesian. It follows that $s', t'$ are étale. It is a formal consequence of the above that $(t', s') : W \to V \times _ S V$ is a monomorphism. We omit the verification that $W \to V \times _ S V$ is an equivalence relation (hint: think about the pullback of $\mathcal{A}$ to $U \times _ X U \times _ X U = R \times _{s, U, t} R$). The quotient sheaf $Y = V/W$ is an algebraic space, see Spaces, Theorem 64.10.5. By Groupoids, Lemma 39.20.7 we see that $Y \times _ X U \cong V$. Hence $Y \to X$ is affine by Lemma 66.20.3. Finally, the isomorphism of

is compatible with glueing isomorphisms, whence $(Y \to X)_*\mathcal{O}_ Y \cong \mathcal{A}$ by Properties of Spaces, Proposition 65.32.1. We omit the verification that this construction is compatible with base change. $\square$

Definition 66.20.8. Let $S$ be a scheme. Let $X$ be an algebraic space over $S$. Let $\mathcal{A}$ be a quasi-coherent sheaf of $\mathcal{O}_ X$-algebras. The *relative spectrum of $\mathcal{A}$ over $X$*, or simply the *spectrum of $\mathcal{A}$ over $X$* is the affine morphism $\underline{\mathop{\mathrm{Spec}}}(\mathcal{A}) \to X$ corresponding to $\mathcal{A}$ under the equivalence of categories of Lemma 66.20.7.

Forming the relative spectrum commutes with arbitrary base change.

Remark 66.20.9. Let $S$ be a scheme. Let $f : Y \to X$ be a quasi-compact and quasi-separated morphism of algebraic spaces over $S$. Then $f$ has a canonical factorization

This makes sense because $f_*\mathcal{O}_ Y$ is quasi-coherent by Lemma 66.11.2. The morphism $Y \to \underline{\mathop{\mathrm{Spec}}}_ X(f_*\mathcal{O}_ Y)$ comes from the canonical $\mathcal{O}_ Y$-algebra map $f^*f_*\mathcal{O}_ Y \to \mathcal{O}_ Y$ which corresponds to a canonical morphism $Y \to Y \times _ X \underline{\mathop{\mathrm{Spec}}}_ X(f_*\mathcal{O}_ Y)$ over $Y$ (see Lemma 66.20.7) whence a factorization of $f$ as above.

Lemma 66.20.10. Let $S$ be a scheme. Let $f : Y \to X$ be an affine morphism of algebraic spaces over $S$. Let $\mathcal{A} = f_*\mathcal{O}_ Y$. The functor $\mathcal{F} \mapsto f_*\mathcal{F}$ induces an equivalence of categories

Moreover, an $\mathcal{A}$-module is quasi-coherent as an $\mathcal{O}_ X$-module if and only if it is quasi-coherent as an $\mathcal{A}$-module.

**Proof.**
Omitted.
$\square$

Lemma 66.20.11. Let $S$ be a scheme. Let $B$ be an algebraic space over $S$. Suppose $g : X \to Y$ is a morphism of algebraic spaces over $B$.

If $X$ is affine over $B$ and $\Delta : Y \to Y \times _ B Y$ is affine, then $g$ is affine.

If $X$ is affine over $B$ and $Y$ is separated over $B$, then $g$ is affine.

A morphism from an affine scheme to an algebraic space with affine diagonal over $\mathbf{Z}$ (as in Properties of Spaces, Definition 65.3.1) is affine.

A morphism from an affine scheme to a separated algebraic space is affine.

**Proof.**
Proof of (1). The base change $X \times _ B Y \to Y$ is affine by Lemma 66.20.5. The morphism $(1, g) : X \to X \times _ B Y$ is the base change of $Y \to Y \times _ B Y$ by the morphism $X \times _ B Y \to Y \times _ B Y$. Hence it is affine by Lemma 66.20.5. The composition of affine morphisms is affine (see Lemma 66.20.4) and (1) follows. Part (2) follows from (1) as a closed immersion is affine (see Lemma 66.20.6) and $Y/B$ separated means $\Delta $ is a closed immersion. Parts (3) and (4) are special cases of (1) and (2).
$\square$

Lemma 66.20.12. Let $S$ be a scheme. Let $X$ be a quasi-separated algebraic space over $S$. Let $A$ be an Artinian ring. Any morphism $\mathop{\mathrm{Spec}}(A) \to X$ is affine.

**Proof.**
Let $U \to X$ be an étale morphism with $U$ affine. To prove the lemma we have to show that $\mathop{\mathrm{Spec}}(A) \times _ X U$ is affine, see Lemma 66.20.3. Since $X$ is quasi-separated the scheme $\mathop{\mathrm{Spec}}(A) \times _ X U$ is quasi-compact. Moreover, the projection morphism $\mathop{\mathrm{Spec}}(A) \times _ X U \to \mathop{\mathrm{Spec}}(A)$ is étale. Hence this morphism has finite discrete fibers and moreover the topology on $\mathop{\mathrm{Spec}}(A)$ is discrete. Thus $\mathop{\mathrm{Spec}}(A) \times _ X U$ is a scheme whose underlying topological space is a finite discrete set. We are done by Schemes, Lemma 26.11.8.
$\square$

## Post a comment

Your email address will not be published. Required fields are marked.

In your comment you can use Markdown and LaTeX style mathematics (enclose it like `$\pi$`

). A preview option is available if you wish to see how it works out (just click on the eye in the toolbar).

Unfortunately JavaScript is disabled in your browser, so the comment preview function will not work.

All contributions are licensed under the GNU Free Documentation License.

## Comments (0)