Let $A$ be a ring, and let $S$ be the scheme $\textrm{Spec }A$. Then the $S$-scheme $G = \textrm{Spec }A[T]$ has the structure of a group scheme over $S$, via the morphisms of $S$-schemes $$m: G \times_S G \rightarrow G, \iota: G \rightarrow G, e: S \rightarrow G$$

whose corresponding $A$-algebra homomorphisms are given respectively by

$$m^{\ast}: A[T] \rightarrow A[X,Y], \space \space f(T) \mapsto f(X+Y)$$ $$\iota^{\ast}: A[T] \rightarrow A[T], \space \space f(T) \mapsto f(-T)$$

$$e^{\ast}: A[T] \rightarrow A, \space \space f(T) \mapsto f(0)$$

If $X$ is any scheme over $S$, then we can identify $$G(X) = \textrm{Hom}_{S-\textrm{sch}}(X,G) = \textrm{Hom}_{A-\textrm{alg}}(A[T], \mathcal O_X(X)) = \mathcal O_X(X)$$

and the morphisms endow $G(X) = \mathcal O_X(X)$ with a group structure identical to the existing additive group structure on $\mathcal O_X(X)$.

Now, what if we replace the affine scheme $S$ with an arbitrary scheme? Does there exist an $S$-group scheme $G$ such that $G(X) = (\mathcal O_X(X),+)$ for all $S$-schemes $X$? The first thing I thought to try was $\textrm{Spec } \mathcal O_S(S)[T]$, but this probably isn't the correct object.

  • 1
    $\begingroup$ Don't know why this was downvoted, +1 $\endgroup$ – user223391 Jun 5 '17 at 22:17

As I typed my question, I realized the correct object to look at was

$$G = S \times \textrm{Spec } \mathbb{Z}[T]$$

(product in the category of schemes). For an $S$-scheme $X$, to give a morphism of $S$-schemes $X \rightarrow G$ is the same as giving a morphism of schemes $X \rightarrow \textrm{Spec } \mathbb{Z}[T]$, so we can identify

$$G(X) = \textrm{Hom}_{\textrm{Sch}}(X, \textrm{Spec } \mathbb{Z}[T]) = \textrm{Hom}_{\textrm{Ring}}(\mathbb{Z}[T], \mathcal O_X(X)) = \mathcal O_X(X)$$

The group operations on $G$ are obtained by taking the obvious group operations on $\textrm{Spec } \mathbb{Z}[T]$, and tensoring with $1_S$.

  • $\begingroup$ Another way to think about this is to use the general fact that if $H$ is a group scheme over $S_0$, and $S \rightarrow S_0$ is a morphism of schemes, then $H \times_{S_0} S$ is a group scheme over $S$. Now take $H = \textrm{Spec } \mathbb{Z}[T], S_0 = \textrm{Spec } \mathbb{Z}$. $\endgroup$ – D_S Jun 5 '17 at 20:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.