Suppose $X$ is a scheme and $f_1,...,f_n\in \Gamma(X,\mathcal O)$ are global sections.
One often reads about the ideal sheaf $\mathcal I=\mathcal (f_1,...,f_n)\subset \mathcal O$, but I have never seen it defined.
The definition should be that $\mathcal I$ is the sheaf associated to the presheaf $\mathcal I^-$ whose value on an open subset $U\subset X$ is $\mathcal I^-(U)=({f_1}{\vert_U},...,{f_n}{\vert_U})\cdot\mathcal O(U)$. Right?
My question is simply: Is there a non-tautological sufficient condition on $U$ and the $f_i$'s guaranteeing that $\mathcal I(U)=\mathcal I^-(U)$?
The only sufficient condition I can think of is that $n=1$ and $U$ is an arbitrary open subset of an integral scheme $X$.

  • 5
    $\begingroup$ A sufficient condition is that $U$ is affine. Indeed $\mathcal{I}$ fits in a s.e.s. $0\to \mathcal{K}\to \mathcal{O}^n\to \mathcal{I}\to 0$ and $H^1(U,\mathcal{K}) = 0$ for $U$ affine. $\endgroup$
    – Chris
    Jul 13, 2020 at 16:08
  • 1
    $\begingroup$ @Christ: Thank you very much for this fine answer: +1. Why not transform this comment in a genuine answer? $\endgroup$
    – lefuneste
    Jul 14, 2020 at 11:26

1 Answer 1


@Chris's comment, posted as an answer by request (made CW to avoid reputation):

A sufficient condition is that $U$ is affine. Indeed $\mathcal I$ fits in an s.e.s. $0 \to \mathcal K \to \mathcal O^n \to \mathcal I \to 0$ and $H^1(U, \mathcal K) = 0$ for $U$ affine.


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.