Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

Let $X$ be a normal projective variety and $D$ a Cartier divisor on it. A point $p\in X$ is called a fixed point of $|D|$ if $p \in \operatorname{supp}(D')$ for any $D'\in |D|$. Here $|D|$ is the linear system of $D$ defined by $|D|=(H^0(X,O(D))\setminus0)/\mathbb{C}^\times\cong \mathbb{P}^{h^0(X,O(D))-1}$.

Could anyone give me a simple example of $X$ and $D$, where the fixed point set of $|D|$ is non-empty?

share|improve this question
add comment

2 Answers

up vote 1 down vote accepted

If $X$ is a smooth projective curve (e.g. a Riemann surface if you work over $\textbf C$) and $D$ is the divisor of a point $P\in X$, then $|D|$ is just $P$ (unless $X\cong\textbf P^1$, in which case $|D|=\textbf P^1$).

As another example, consider this pencil ($=$ one-dimensional linear system) on the projective plane: for $t$ ranging $\textbf P^1$, the corresponding divisor is given by the equation $y(x-tz)$: the $x$-axis plus a "vertical line". All these divisors meet at the point at infinity, and there is no other base point (even if the $x$-axis appears in each divisor!)

A useful characterization says that $|D|$ is base-point-free if and only if $\mathscr O_X(D)$ is generated by its global sections. Similarly, for non-complete linear systems: $\textbf PV^\ast\subseteq |D|$ is base-point-free if and only if $\mathscr O_X(D)$ is generated by the global sections in (a basis of) $V$. Translation: a linear system determines a rational map $\psi:X\dashrightarrow \textbf PV^\ast$, and $x\in X$ is a base point if and only if $\psi$ is not defined at $x$ (which means that all the sections generating $V$ vanish at $x$). Just take $V=H^0(X,\mathscr O_X(D))$ if you only care about complete linear systems.

Another example of a linear system with base points: let $k$ be the algebraic closure of $\textbf F_2$ and let $V\subset H^0(\textbf P^2_k,\mathscr O(3))$ be spanned by the cubics

\begin{equation} x^2y+xy^2,\,x^2z+z^2x,\,y^2z+z^2y. \end{equation}

Then the linear system $\textbf PV^\ast$ (which is two-dimensional) has $7$ base points: look at those points in $\textbf P^2_k$ at which the above sections vanish simultaneously, and you will find that this happens exactly when the coordinates all lie in $\textbf F_2$. So the base locus here is the finite projective plane $\textbf P^2_{\textbf F_2}$.

share|improve this answer
    
I totally forgot about the useful criterion that $|D|$ is base-point-free iff $O(D)$ is generated by global sections. Then the ideal sheaf $I_{Z}=O(Z)$ of subvariety $Z\subset X$ is not generated by global sections and thus not base=point-free. I don't quite understand your second example. Aren't points on $y=0$ base-point of this linear system? –  M. K. Sep 30 '12 at 22:58
    
@M. K. You are right: if $(x:y:z)$ are homogeneous coordinates on $\textbf P^2$, then the base locus of that linear system is the copy of $\textbf P^1$ given by $y=0$. What I had in mind (but, I'm sorry, I did not write it correctly!) is an example by Kleiman (arxiv.org/pdf/alg-geom/9704018v1.pdf, p. 12), involving inhomogeneous coordinates. –  Brenin Oct 1 '12 at 18:10
    
Thank you for clarifying the example. I will check the Keliman's article too. By the way, in my comments $O(Z)$ is typo and should be $O(-Z)$. –  M. K. Oct 1 '12 at 18:45
add comment
  1. If $D<0$, then every point of $X$ is a fixed point of $|D|=\emptyset$.

  2. If $X$ is a smooth projective curve of genus $\ge 1$, then for any rational point $P$ of $X$, $P$ is a fixed point of $|P|$ because otherwise, there would exits a rational function on $X$ with only a simple pole and this would imply that $X$ is the projective line.

  3. Let $X$ be as a curve as above. Let $P\in X$. Then $|2P|$ has no fixed point if and only if $P$ is a so called Weierstrass point. On a hyperelliptic curve, they are the ramification points of the canonical degree $2$ map $X\to \mathbb P^1$.

  4. Let $f$ be a rational map from $\mathbb P^n$ to $\mathbb P^m$. Let $D$ be the pullback of a hyperplane by $f$. Then $D$ has no fixed points if and only if $f$ is defined everywhere. So take for example f equal to a projection to a linear subspace, then $D$ has fixed points.

share|improve this answer
    
Thank you for providing me with many examples. Just a quick question, is every point of $X$ a fixed point if $|D|=\emptyset$? The second example is a easy good example. –  M. K. Sep 30 '12 at 22:36
    
@M.K.: The first example is kind of cheating. But well logically it is correct because $|D|=\emptyset$ means there is no condition to verifiy for $p\in X$. –  user18119 Oct 1 '12 at 7:45
add comment

Your Answer

 
discard

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.