next up previous
Next: Symbolic Hamburger-Noether expressions Up: Effective Construction of Algebraic Previous: Effective Construction of Algebraic

Preliminaries, Notations

Throughout the following, let $ \mathbf{F}$ be a finite field and $ \overline{\mathbf{F}}$ an algebraic closure of $ \mathbf{F}$. Moreover, let $ C\subset
\mathbb {P}^2(\mathbf{F})$ be an absolutely irreducible, reduced, projective plane curve given by a homogeneous form of degree $ d$, $ F\in \mathbf{F}[X,Y,Z]_d$.

A point $ P\in C$ is called rational if its coordinates are in $ \mathbf{F}$. More generally, by a closed point $ [P]\in C$ we denote the formal sum of a point (defined over $ \overline{\mathbf{F}}$) with its conjugates. If there is no risk of confusion, we sometimes write $ P\in C$ to denote the closed point $ [P]$. Note that closed points are invariant under the action of the Galois group $ Gal(\overline{\mathbf{F}}/\mathbf{F})$.

We denote by $ n:{\widetilde{C}}\to C$ the normalization and by

$\displaystyle N:=i\circ n: \;{\widetilde{C}}\to
\mathbb {P}^2(\mathbf{F})$

the parametrization of $ C$ (with $ i:C\hookrightarrow
\mathbb {P}^2(\mathbf{F})$ being the inclusion).

The points of $ {\widetilde{C}}$ are called places of $ C$. Again, a place is called rational if its coordinates are in $ \mathbf{F}$, and by a closed place we denote the formal sum of a place (defined over $ \overline{\mathbf{F}}$) with its conjugates. Note that each smooth (rational, resp. closed) point $ P\in C$ corresponds to a unique (rational, resp. closed) place $ n^{-1}(P)\in
{\widetilde{C}}$. If $ P$ is a singular point of $ C$ then each local branch of $ C$ at $ P$ corresponds to a unique place of $ C$. Hence, the set of places of $ C$ can be identified with the set consisting of the non-singular points of $ C$ and all tuples $ \bigl(P;\!(C_i,P)\bigr)$, $ P$ a singular point of $ C$ and $ (C_i,P)$ a local branch of $ C$ at $ P$.

A (rational) divisor $ D$ on $ {\widetilde{C}}$ is a finite, weighted, formal sum of (closed) places $ Q$ of $ C$, $ D = \sum n_{Q} Q$ with integer coefficients $ n_Q =: ord_Q (D)$. The divisor $ D$ is called effective if there are no negative $ n_Q$. Moreover, we introduce the degree of the divisor $ D$, $ \deg D:=\sum n_Q$ and the support of $ D$, supp$ \,D:=\{Q \mid n_Q\neq 0\}$.

To each element $ g$ in the function field $ \mathbf{F}({\widetilde{C}})$ of $ {\widetilde{C}}$ one associates the principal divisor $ (g):= \sum ord_Q(g)\cdot Q$. Note that $ (g)$ has degree 0, by the residue theorem.

Finally, the linear system associated to a divisor $ D$ is defined to be

$\displaystyle \mathcal{L}(D)\,:=\, \bigl\{\, g\in \mathbf{F}({\widetilde{C}}) \, \mid \, (g)\geq
-D\, \bigr\} \cup \bigl\{ 0 \bigr\}\,,$

where the order relation is defined by

\begin{displaymath}D\geq D' \;:\Longleftrightarrow \;\, ord_Q(D)\geq ord_Q(D')
\,\text{ for all places $Q\in {\widetilde{C}}$}\,.\end{displaymath}


next up previous
Next: Symbolic Hamburger-Noether expressions Up: Effective Construction of Algebraic Previous: Effective Construction of Algebraic
Christoph Lossen
2001-03-21