Degree formula for connective K-theory.pdf

a r X i v : 0 8 0 8 .2 5 9 2 v 2 [ m a t h .A G ] 1 5 S e p 2 0 0 8 Degree formula for connective K-theory K. Zainoulline September 15, 2008 Abstract We apply the degree formula for connective K-theory to study incom- pressibility of algebraic varieties. 1 1 Introduction The celebrated Rost degree formula says that given a rational map f : Y 99K X between two smooth projective varieties there is the congruence relation (see [9]) ηp(Y ) ≡ deg f · ηp(X) mod nX , (1) where p is a prime, ηp(X) is the Rost number of X , deg f is the degree of f and nX is the greatest common divisor of degrees of all closed points on X (see [6], [7], [8] and [9]). It was conjectured by Rost that the degree formula (1) should follow from a generalized degree formula for some universal cohomology theory. This conjec- ture was proven by Levine and Morel in [5], where they constructed the theory of algebraic cobordism ? and provided the respective degree formula (see [5, Theorem 1.2.14]). Unfortunately, the generalized degree formula has one disadvantage: it deals with elements in the cobordism ring which is too big and usually is hard to compute. On the other hand the classical degree formula (1) is easy to apply but it catches only “pro-p” effects. The reasonable question would be to find a cohomology theory together with a degree formula which doesn’t loose much information and is still computable. The natural candidate for such a theory is the connective K-theory denoted by K. It has two important properties: First, K is the universal oriented co- homology theory for the Chow group CH and Grothendieck’s K0, meaning the following diagram of natural transformations ? prK  CH K //oo K0[β, β?1], 1 MSC: 14F43; 55N20; 55N22 Keywords: connective K-theory, degree formula 1 where β denotes the Bott element. Second, it is the universal birational theory in the sense that it preserves the fundamental classes for birational maps, i.e. for any proper birational f : Y → X we have f?(1Y ) = 1X