Let $F(x_0, x_1, x_2)$ be a homogeneous polynomial with $deg( f) = 2$. Let $V:= Z(F) \in \mathbb{P}^2$ be the zero-set of $F$.

The task is to show that $V \simeq \mathbb{P}^1$.

Firstly, we can change co-ordinates such that $F = X_0^2 - X_1X_2$. After some discussion with a friend, I realised that, by employing the morphism,

$\phi: \mathbb{P}^1 \to V \\ \phi: (u;t) \mapsto (u^2, t^2, ut)$

and then considering the restriction of $\phi$ to the each open set in the standard affine open cover of $\mathbb{P}^1$, we can show that the map is well-defined, and indeed an isomorphism of varieties.

My question is this: is it possible to establish the same result, but instead, by establishing a $k$-algebra isomorphism of the underlying co-ordinate rings? Indeed, what really are the co-ordinate rings of projective varieties?

Moreover, does the following argument hold?

Consider the k-algebra homomorphism induced by,

$\psi: k[x_0,x_1,x_2] \to k[u,t] \\ x_0 \mapsto u^2 \\ x_1 \mapsto t^2 \\ x_2 \mapsto ut$

Then, we have $ker (\psi) = (F)$ and so, $Im(\psi) \simeq \frac{k[x_0,x_1,x_2]}{F} = A(V)$ by the isomorphism theorem and the definition of the co-ordinate ring of V.

But then, $Im(\psi) = k[u^2, t^2, ut]$, which is the co-ordinate ring of $\mathbb{P}^1$.

Therefore, as $A(V) \simeq A(\mathbb{P}^1)$, we have $V \simeq \mathbb{P}^1$.


1) No, it is not possible to
"establish the same result...by establishing a $k$-algebra isomorphism of the underlying co-ordinate rings"
because these coordinate rings are not isomorphic!
Indeed the ring $k[x_0,x_1,x_2]:=\frac{k[X_0,X_1,X_2]}{(X_0^2-X_1X_2)}$ is not isomorphic to $k[T_0,T_1]$ since the first is not a UFD ($x_0^2=x_1x_2$ are different factorizations of the same element!), whereas the second is a UFD, like all polynomial rings over a field.

2) The notion of "coordinate ring" for a projective variety $X$ depends on the embedding of $X$ into projective space and is by far not as satisfying as in the affine case.
In particular that coordinate ring $S(X)$ is not at all equal to the ring of regular functions $\mathcal O(X)=k$.

3) It is not true that "we can change co-ordinates such that $F = X_0^2 - X_1X_2$".
This is valid only for irreducible polynomials $F$.
For example no change of coordinates will force the conic $X_0X_1=0$ to have as equation $X_0^2 - X_1X_2$ .


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.