The Stacks project

77.20 Quotient stacks

In this section and the next few sections we describe a kind of generalization of Section 77.19 above and Groupoids, Section 39.20. It is different in the following way: We are going to take quotient stacks instead of quotient sheaves.

Let us assume we have a scheme $S$, and algebraic space $B$ over $S$ and a groupoid in algebraic spaces $(U, R, s, t, c)$ over $B$. Given these data we consider the functor
\begin{equation} \label{spaces-groupoids-equation-quotient-stack} \begin{matrix} (\mathit{Sch}/S)_{fppf}^{opp} & \longrightarrow & \textit{Groupoids} \\ S' & \longmapsto & (U(S'), R(S'), s, t, c) \end{matrix} \end{equation}

By Categories, Example 4.37.1 this “presheaf in groupoids” corresponds to a category fibred in groupoids over $(\mathit{Sch}/S)_{fppf}$. In this chapter we will denote this

\[ [U/_{\! p}R] \to (\mathit{Sch}/S)_{fppf} \]

where the subscript ${}_ p$ is there to distinguish from the quotient stack.

Definition 77.20.1. Quotient stacks. Let $B \to S$ be as above.

  1. Let $(U, R, s, t, c)$ be a groupoid in algebraic spaces over $B$. The quotient stack

    \[ p : [U/R] \longrightarrow (\mathit{Sch}/S)_{fppf} \]

    of $(U, R, s, t, c)$ is the stackification (see Stacks, Lemma 8.9.1) of the category fibred in groupoids $[U/_{\! p}R]$ over $(\mathit{Sch}/S)_{fppf}$ associated to (

  2. Let $(G, m)$ be a group algebraic space over $B$. Let $a : G \times _ B X \to X$ be an action of $G$ on an algebraic space over $B$. The quotient stack

    \[ p : [X/G] \longrightarrow (\mathit{Sch}/S)_{fppf} \]

    is the quotient stack associated to the groupoid in algebraic spaces $(X, G \times _ B X, s, t, c)$ over $B$ of Lemma 77.15.1.

Thus $[U/R]$ and $[X/G]$ are stacks in groupoids over $(\mathit{Sch}/S)_{fppf}$. These stacks will be very important later on and hence it makes sense to give a detailed description. Recall that given an algebraic space $X$ over $S$ we use the notation $\mathcal{S}_ X \to (\mathit{Sch}/S)_{fppf}$ to denote the stack in sets associated to the sheaf $X$, see Categories, Lemma 4.38.6 and Stacks, Lemma 8.6.2.

Lemma 77.20.2. Assume $B \to S$ and $(U, R, s, t, c)$ as in Definition 77.20.1 (1). There are canonical $1$-morphisms $\pi : \mathcal{S}_ U \to [U/R]$, and $[U/R] \to \mathcal{S}_ B$ of stacks in groupoids over $(\mathit{Sch}/S)_{fppf}$. The composition $\mathcal{S}_ U \to \mathcal{S}_ B$ is the $1$-morphism associated to the structure morphism $U \to B$.

Proof. During this proof let us denote $[U/_{\! p}R]$ the category fibred in groupoids associated to the presheaf in groupoids ( By construction of the stackification there is a $1$-morphism $[U/_{\! p}R] \to [U/R]$. The $1$-morphism $\mathcal{S}_ U \to [U/R]$ is simply the composition $\mathcal{S}_ U \to [U/_{\! p}R] \to [U/R]$, where the first arrow associates to the scheme $S'/S$ and morphism $x : S' \to U$ over $S$ the object $x \in U(S')$ of the fibre category of $[U/_{\! p}R]$ over $S'$.

To construct the $1$-morphism $[U/R] \to \mathcal{S}_ B$ it is enough to construct the $1$-morphism $[U/_{\! p}R] \to \mathcal{S}_ B$, see Stacks, Lemma 8.9.2. On objects over $S'/S$ we just use the map

\[ U(S') \longrightarrow B(S') \]

coming from the structure morphism $U \to B$. And clearly, if $a \in R(S')$ is an “arrow” with source $s(a) \in U(S')$ and target $t(a) \in U(S')$, then since $s$ and $t$ are morphisms over $B$ these both map to the same element $\overline{a}$ of $B(S')$. Hence we can map an arrow $a \in R(S')$ to the identity morphism of $\overline{a}$. (This is good because the fibre category $(\mathcal{S}_ B)_{S'}$ only contains identities.) We omit the verification that this rule is compatible with pullback on these split fibred categories, and hence defines a $1$-morphism $[U/_{\! p}R] \to \mathcal{S}_ B$ as desired.

We omit the verification of the last statement. $\square$

Lemma 77.20.3. Assumptions and notation as in Lemma 77.20.2. There exists a canonical $2$-morphism $\alpha : \pi \circ s \to \pi \circ t$ making the diagram

\[ \xymatrix{ \mathcal{S}_ R \ar[r]_ s \ar[d]_ t & \mathcal{S}_ U \ar[d]^\pi \\ \mathcal{S}_ U \ar[r]^-\pi & [U/R] } \]


Proof. Let $S'$ be a scheme over $S$. Let $r : S' \to R$ be a morphism over $S$. Then $r \in R(S')$ is an isomorphism between the objects $s \circ r, t \circ r \in U(S')$. Moreover, this construction is compatible with pullbacks. This gives a canonical $2$-morphism $\alpha _ p : \pi _ p \circ s \to \pi _ p \circ t$ where $\pi _ p : \mathcal{S}_ U \to [U/_{\! p}R]$ is as in the proof of Lemma 77.20.2. Thus even the diagram

\[ \xymatrix{ \mathcal{S}_ R \ar[r]_ s \ar[d]^ t & \mathcal{S}_ U \ar[d]^{\pi _ p} \\ \mathcal{S}_ U \ar[r]^-{\pi _ p} & [U/_{\! p}R] } \]

is $2$-commutative. Thus a fortiori the diagram of the lemma is $2$-commutative. $\square$

Remark 77.20.4. In future chapters we will use the ambiguous notation where instead of writing $\mathcal{S}_ X$ for the stack in sets associated to $X$ we simply write $X$. Using this notation the diagram of Lemma 77.20.3 becomes the familiar diagram

\[ \xymatrix{ R \ar[r]_ s \ar[d]_ t & U \ar[d]^\pi \\ U \ar[r]^-\pi & [U/R] } \]

In the following sections we will show that this diagram has many good properties. In particular we will show that it is a $2$-fibre product (Section 77.22) and that it is close to being a $2$-coequalizer of $s$ and $t$ (Section 77.23).

Comments (2)

Comment #1812 by OS on

In \label{equation-quotient-stack} I guess it should not be but . There are several other places below where the same change would be necessary.

Comment #1830 by on

Hmm.... no! The point is that we always want our categories fibred in groupoids to be on the category of schemes over . What you are sayng is actually discussed in Lemma 77.20.2.

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.

In order to prevent bots from posting comments, we would like you to prove that you are human. You can do this by filling in the name of the current tag in the following input field. As a reminder, this is tag 044O. Beware of the difference between the letter 'O' and the digit '0'.