The first day of Algebraic Curves taught by Karl Christ. We briefly discussed the course structure and Karl gave a brief overview of the type of math we can expect to see.
Instructor: Karl Christ
References:
Harshorne Chapter IV
Arbasello, Corualla, Griffiths, Harris: Geometry of Algebraic Curves, Vol I (maybe Vol II also)
Eisenbud and Harris: The practice of algebraic curves (Draft)
Haris: Basic algebraic geometry for constructions of things like in terms of coordinates, where you can really get your hands dirty.
Grade: There is no regular homework. There is a final exam, which is an oral exam. You can get extra credit by handing in up to 2 papers about 5-10 pages each. Topic can be anything from the class.
Office Hours: 15:30 - 17:00 on Monday. Can also talk after class on Wednesday at department coffee.
The most natural (first) way to study algebraic curves is by interpreting them as the zero sets of polynomials. These are embedded algebraic curves. These have been studied almost since the inception of mathematics, the Ancient Greeks studied them for instance. We will typically interpret embedded algebraic curves as subsets of or more concretely in .
We will also study abstract curves i.e. one-dimensional schemes i.e. compact 1-dimensional complex manifolds (Riemann surfaces).
The main theme of this course will be the study of an algebraic curve together with a map . This allows us to study both abstract and embedded curves simultaneously.
Perhaps the first thing to not is that this adds nothing new to the theory.
Note that this fails for singular curves; the dimension of the Zariski tangent space can be arbitrarily large at a singularity and any embedding of a curve must map to a space whose dimension is at least as large as this dimension. You can find singularities whose Zariski tangent space is of arbitrarily large dimension.
The map is given by a line bundle on and a vector subspace of of dimension , where is the degree of .
We thus far have two invariants: the dimension of the ambient projective space and the degree of the embedding . The third primary invariant we consider is the genus of . A natural question is this: given and , what are the possible values of ?
Definition: Let be a curve.
The arithmetic genus of is where is the Hilbert polynomial of .
The geometric genus of is where is the canonical sheaf .
These two notions agree for curves and are equal to by Serre duality. We therefore simply refer to the genus of and denote it .
For we have the degree-genus formula: .
For it is no longer true that depends solely on .
Castelnnovo's bound gives an upper bound to
A classification was given by Hartshorne.
For any there is a generalization of Castelnnovo's bound, but giving the possible values is an open problem.
Let's instead ask a different question: given fixed and , what are the possible values of ?
Riemann-Roch: , which implies . This gives a lower bound on .
Clifford Theorem: if . This gives an upper bound on .
When , the lower bound and upper bound of coincide, and then fully determines . For however, the upper and lower bounds for do not coincide giving a "special region of line bundles". This region is exactly what the Brill Noether theorem describes.
There's an even stronger theorem: