0
$\begingroup$

Let $f:X \rightarrow Y$ be a dominant morphisms between two irreducible algebraic affine complex varieties. If $X$ and $Y$ have the same dimension, it is known that in this case we have two very nice properties: (i) the map $y \mapsto |f^{-1}(y)|$ is constant over a non-empty open (with respect to the Zariski topology on $Y$) subset of $Y$, and (ii) every regular value of $Y$ is a local minimum of the map $y \mapsto |f^{-1}(y)|$.

My question is the following: What other types of properties exist that behave similarly for dominant morphisms?

For a particular example I am interested in, now allow $\dim X \geq \dim Y$ in our original set up. Assuming $y \in Y$ is a regular value of $f$, we can ask what the genus of the algebraic set $f^{-1}(y)$ is. Is it true in this case that the map $y \mapsto \text{genus of } f^{-1}(y)$ is (i) constant over a non-empty subset of $Y$, and (ii) locally minimised/maximised at every regular value of $Y$? What other similar such properties are there (e.g., degree, number of irreducible components, etc.)?

Any references to back up any claims will be particularly helpful. Since my knowledge regarding algebraic geometry is rather limited to the basics, any answering utilising schemes will most likely be useless for my purposes.

$\endgroup$
3
  • 1
    $\begingroup$ Not really sure what notion of genus you are referring to (I know what that means for smooth projective curves), but some examples are projection $\mathbf C^2 \setminus \{0\} \to \mathbf C$ (most fibres are $\mathbf C$ but one fibre is $\mathbf C \setminus \{0\}$) and projection $\{(x,y) \in \mathbf C^2\ |\ xy\neq 1\} \to \mathbf C$ (most fibres are $\mathbf C \setminus \{0\}$ but one fibre is $\mathbf C$). So the first Betti number $\dim H_1(f^{-1}(y),\mathbf Q)$ can jump up or down. $\endgroup$ Commented Jun 9, 2023 at 17:45
  • 1
    $\begingroup$ In general, a lot of numerical quantities $N$ of $f^{-1}(y)$ are constructible in $y$, i.e. the sets $Y_n = \{y \in Y\ |\ N(f^{-1}(y)) = n\}$ are constructible. A lot of these properties are collected in EGA IV$_3$, §9 (although I'm always disappointed that the one I'm looking for is never there...). $\endgroup$ Commented Jun 9, 2023 at 17:52
  • $\begingroup$ 1) You would probably want to assume that $f$ is proper to get anything like what you are hoping for. 2) In any case, all morphisms are generally flat. In other words, there is an open subset of $Y$ over which $f$ is flat. A lot of those things you are considering follow from this (if you assume that $f$ is proper...) $\endgroup$ Commented Jun 12, 2023 at 5:42

0

You must log in to answer this question.

Browse other questions tagged .