This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: PL ¨ UCKER FORMULAS Let f : X → P n be a map of a nonsingular projective curve X . It is given by an invertible sheaf L of some degree d and a basis ( s , . . . , s n ) of a linear subspace V of H ( X, L ). In coordinatefree way, we have a linear subspace V ⊂ H ( X, L ) of dimension n + 1 and the map X → P ( V * ) defined by f ( x )( s ) = P ( { s : s ( x ) = 0 } ) for any s ∈ V . For any point x ∈ X and s ∈ V \{ } we denote by ν x ( s ) the largest power of the maximal ideal m i X,x such that s x ∈ m i X,x L x . Let V i = ν 1 x ( Z ≥ i ) , i ≥ 0. Since dim m i X,x / m i +1 X,x = 1, the quotient space V i /V i +1 is of dimension ≤ 1. This implies that there is a unique sequence of nonnegative numbers 0 ≤ α 1 ≤ . . . ≤ α n such that (1) { } ⊂ V n + α n ⊂ . . . ⊂ V 1+ α 1 ⊂ V, where codim V i + α i = i. The linear projective subspace P O i 1 ( f, x ) = P ( V ⊥ i + α i ) of P ( V * ) of dimension i 1 is called the osculating (i1)plane of f at the point x . The sequence of projective subspaces (2) P O ( f, x ) ⊂ . . . ⊂ P O n 1 ( f, x ) ⊂ P ( V * ) is called the osculating flag of f at x with ramification sequence ( α 1 , . . . , α n ). Here the subscript indicates the dimension of the subspace. It is clear that P O ( f, x ) = f ( x ). Let us choose projective coordinates in such a way that (3) P O i ( f, x ) = { T i +1 = . . . = T n = 0 } , i = 0 , . . . , n 1 . Let t i = T i /T be inhomogeneous coordinates, and let t be a local coordinate of X at x , the map f is given locally by (4) ( x 1 , . . . , x n ) = ( t 1+ α 1 g 1 ( t ) , t 2+ α 2 g 2 ( t ) , . . . , t n + α n g n ( t )) , where g i (0) 6 = 0. Under the homomorphism of ring k [ t 1 , . . . , t n ] → O P ( V * ) ,f ( x ) → O X,x the images of the functions t i belong to m i + α i X,x . In particular, if α 1 > 0, all functions belong to m 2 X,x and hence the map f is ramified at x . If α 1 = 0, and f 1 ( f ( x )) = { x } , then f is a closed embedding in an affine neighborhood of x . In this case, the osculating hyperplane P O n 1 ( f, x ) intersects f ( X ) at x with multiplicity n + α n . A point x is called a hyperosculating point if α n > 0. Clearly, all ramification points are hyperosculating points. An honest osculating point is a hyperosculating point with α 1 = . . . = α n 1 = 0. Example 1. Assume n = 2 and let C ⊂ P 2 be an embedded plane curve with affine equation F ( x, y ) = 0 , F (0 , 0) = 0 . Let X be its normalization and f : X → C , → P 2 be the composition such that f ( x ) = (0 , 0) . Choose 1 2 PL ¨ UCKER FORMULAS the system coordinates as in (3) . Then the coordinate line x = 0 intersects the curve with multiplicity 1 + α 1 at the origin and the coordinate line y = 0 intersects the curve with multiplicity 2+ α 2 . An honest hyperosculating point with is called an inflection points (of order α n ). Obviously, C is singular at (0 , 0) if α 1 > 1 . If (0 , 0) is an ordinary cusp, then the osculating flag is...
View
Full Document
 Fall '08
 IgorDolgache
 Linear Algebra, Algebra, Geometry, Vector Space, Formulas, pn, Duality, Algebraic curve, Linear subspace

Click to edit the document details