6
$\begingroup$

A complex weighted projective is $\mathbb{P}(k_1, \cdots, k_{n+1})=Proj(\mathbb{C}[x_1, \cdots, x_{n+1}])$ with $x_i$ of degree $k_i$ (sometimes people ask for each $n$ of the weights being coprime). My first question is whether all weighted projective spaces are rational.

The rationality of $\mathbb{P}(k_1, \cdots, k_{n+1})$ is equivalent to the functional field $\mathbb{C}_0(x_1, \cdots, x_{n+1})$ being free generated by $n$ elements. When $n=1$, it is clear that $\mathbb{C}_0(x_1, x_2)=\mathbb{C}(\frac{x_1^{k_2}}{x_2^{k_1}})$, hence $\mathbb{P}(k_1, k_2)$ is always a rational curve. However, I guess in higher dimension this could not hold.

If the answer to the first question is no, then my second question is that whether some weaker rationality properties, like rational connectivity, stable rationality, etc hold for weighted projective space.

Thank you in advance!

$\endgroup$

1 Answer 1

10
$\begingroup$

The $n$-dimensional weighted projective space $X = \mathbb{P}(k_1,\dots,k_{n+1})$ is a toric variety, hence is automatically rational since it contains an isomorphic copy of $\mathbb{G}_m^n$ as a dense open subvariety. Indeed, let $T$ be the torus $\{(t_1,\dots,t_{n+1}) \mid t_i \in \mathbb{G}_m\}/\{(t,\dots,t)\mid t\in \mathbb{G}_m \}$. Then $T$ acts on $X$ via $$(t_1,\dots,t_{n+1}) \cdot [x_1 : \dots :x_{n+1}] = \left[t_1^{k_1}x_1 :\dots :t_{n+1}^{k_{n+1}}x_{n+1} \right]$$ and the orbit of $[1: \dots : 1]$ is isomorphic to $T$, since its stabilizer is a product of cyclic groups.

$\endgroup$
2
  • $\begingroup$ Thank you! The fact that weighted projective space is toric is useful to me. I think there is a small problem with your argument: the kernel of the action is not trivial, it is the product of some finite cyclic groups. But that's fine because the quotient of $\mathbb{G}_m$ by $\mu_k$ is isomorphic to $\mathbb{G}_m$. $\endgroup$ Nov 18, 2019 at 22:32
  • $\begingroup$ Thanks! I have edited the answer $\endgroup$
    – Jef
    Nov 24, 2019 at 11:33

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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