MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

As we know there are a lot of principle of connectedness in algebraic geometry. Here is a useful and interesting one:

Suppose T is an integral curve over k. X-->T is a flat family of closed subvarieties in $P_k^n$. If there is a non-empty open subset U in T such that at every closed point t in U, the fiber X_t is connected. Then show every fiber X_t is connected for any t in T.

In consideration of uppercontinuous property, this says that if the parameter space is a curve, then if $h^0(X_t, O_{X_t})$ is locally constant on some open set, then it's locally constant everywhere!(If we further require k is algebraically closed here).

I thing this property is interesting and useful, but I can't prove it, and every reference I can find traces back to Hartshorne's exercise III.11.4. If anyone can give me a proof, I would be very grateful!

share|cite|improve this question

EDIT: This is a small correction on the previous argument that I posted here, prompted by a comment of Steven Sam.

The question is stable under any base change. Let $\nu : \widetilde{T} \to T$ be the normalization of $T$. Let $\tilde{f} : \widetilde{X} \to \widetilde{T}$ be the base change of $f : X \to T$ via $\nu$. Then $\tilde{f}$ is again flat and projective and $\widetilde{T}$ is smooth. The sheaf $\tilde{f}\_* \mathcal{O}\_{\widetilde{X}}$ is torsion free since $\tilde{f} : \widetilde{X} \to \widetilde{T}$ is flat and projective. Since $\widetilde{T}$ is a smooth curve this implies that $\tilde{f}\_*\mathcal{O}\_{\widetilde{X}}$ is locally free. Since $f$ is assumed to have connected fibers over $U$, we get that $\tilde{f}$ has connected fibers on $U$ and so $\tilde{f}\_* \mathcal{O}\_{\widetilde{X}}$ has rank one on $\nu^{-1}(U)$ and so it must have rank one everywhere. Thus $\tilde{f}\_*\mathcal{O}\_{\widetilde{X}}$ is the sheaf of sections of a line bundle on $\widetilde{T}$. On the other hand the space of global sections of $\tilde{f}\_* \mathcal{O}\_{\widetilde{X}}$ is equal to the space of global sections of $\mathcal{O}\_{\widetilde{X}}$ and so $\tilde{f}\_* \mathcal{O}\_{\widetilde{X}}$ has a nowhere vanishing section corresponding to the section $1$ of $\mathcal{O}\_{\widetilde{X}}$. This shows that $\tilde{f}\_* \mathcal{O}\_{\widetilde{X}}$ is the trivial line bundle, i.e. $\tilde{f}\_* \mathcal{O}\_{\widetilde{X}} \cong \mathcal{O}\_{\widetilde{T}}$. Now this implies that all fibers of $\tilde{f}$ are connected and hence the fibers of $f$ are connected.

share|cite|improve this answer
I took the liberty of fixing your latex: I wanted to read it :P – Mariano Suárez-Alvarez Dec 16 '09 at 5:14
Thanks! I couldn't figure out what I was doing wrong. It is good to know how to fix those subscripts. – Tony Pantev Dec 16 '09 at 5:18
I am not sure why the connectedness of generic fiber can imply that $f_* \mathcal{O}_X$ has rank one generically. Consider this example, take A to be a DVR, hence spec A only has one closed point and one generic point. If we choose B to be a finite extension of A such that the unique prime of A ramifies in B, then Spec B-->Spec A has generic fiber but don't have rank 1. Hence this proposition seems fail to hold for finite proper morphisms. Projectivity may play a more important role rather than just implying $f_* \mathcal{O}_X$ is a coherent sheaf on T, I think. – Taisong Jing Dec 16 '09 at 5:56
In my last comment, I meant Spec B--> Spec A has CONNECTED generic fiber, sorry for missing that information. – Taisong Jing Dec 16 '09 at 5:58
Taisong, in your setup you assumed that your morphism maps you to an integral curve over a field and that in $T$ you have a non-empty open set $U$ over which all fibers of $f$ arre connected. Restricting f over $U$ gives a projectve morphism with connected fibers so the push-forward of the structure sheaf is the structure sheafd. – Tony Pantev Dec 16 '09 at 13:29

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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