Splitting principle

In mathematics, the splitting principle is a technique used to reduce questions about vector bundles to the case of line bundles. In the theory of vector bundles, one often wishes to simplify computations, say of Chern classes. Often computations are well understood for line bundles and for direct sums of line bundles. In this case the splitting principle can be quite useful.

Theorem — Let [math]\displaystyle{ \xi\colon E\rightarrow X }[/math] be a vector bundle of rank [math]\displaystyle{ n }[/math] over a paracompact space [math]\displaystyle{ X }[/math]. There exists a space [math]\displaystyle{ Y=Fl(E) }[/math], called the flag bundle associated to [math]\displaystyle{ E }[/math], and a map [math]\displaystyle{ p\colon Y\rightarrow X }[/math] such that

1. the induced cohomology homomorphism [math]\displaystyle{ p^*\colon H^*(X)\rightarrow H^*(Y) }[/math] is injective, and
2. the pullback bundle [math]\displaystyle{ p^*\xi\colon p^*E\rightarrow Y }[/math] breaks up as a direct sum of line bundles: [math]\displaystyle{ p^*(E)=L_1\oplus L_2\oplus\cdots\oplus L_n. }[/math]

The theorem above holds for complex vector bundles and integer coefficients or for real vector bundles with [math]\displaystyle{ \mathbb{Z}_2 }[/math] coefficients. In the complex case, the line bundles [math]\displaystyle{ L_i }[/math] or their first characteristic classes are called Chern roots.

The fact that [math]\displaystyle{ p^*\colon H^*(X)\rightarrow H^*(Y) }[/math] is injective means that any equation which holds in [math]\displaystyle{ H^*(Y) }[/math] (say between various Chern classes) also holds in [math]\displaystyle{ H^*(X) }[/math].

The point is that these equations are easier to understand for direct sums of line bundles than for arbitrary vector bundles, so equations should be understood in [math]\displaystyle{ Y }[/math] and then pushed down to [math]\displaystyle{ X }[/math].

Since vector bundles on [math]\displaystyle{ X }[/math] are used to define the K-theory group [math]\displaystyle{ K(X) }[/math], it is important to note that [math]\displaystyle{ p^*\colon K(X)\rightarrow K(Y) }[/math] is also injective for the map [math]\displaystyle{ p }[/math] in the above theorem.

The splitting principle admits many variations. The following, in particular, concerns real vector bundles and their complexifications:

Theorem — Let [math]\displaystyle{ \xi\colon E\rightarrow X }[/math] be a real vector bundle of rank [math]\displaystyle{ 2n }[/math] over a paracompact space [math]\displaystyle{ X }[/math]. There exists a space [math]\displaystyle{ Y }[/math] and a map [math]\displaystyle{ p\colon Y\rightarrow X }[/math] such that

1. the induced cohomology homomorphism [math]\displaystyle{ p^*\colon H^*(X)\rightarrow H^*(Y) }[/math] is injective, and
2. the pullback bundle [math]\displaystyle{ p^*\xi\colon p^*E\rightarrow Y }[/math] breaks up as a direct sum of line bundles and their conjugates: [math]\displaystyle{ p^*(E\otimes \mathbb{C})=L_1\oplus \overline{L_1}\oplus \cdots\oplus L_n\oplus \overline{L_n}. }[/math]

Symmetric polynomial

Under the splitting principle, characteristic classes for complex vector bundles correspond to symmetric polynomials in the first Chern classes of complex line bundles; these are the Chern classes.

See also

• K-theory
• Grothendieck splitting principle for holomorphic vector bundles on the complex projective line

References

• Hatcher, Allen (2003), Vector Bundles & K-Theory (2.0 ed.), http://www.math.cornell.edu/~hatcher/VBKT/VBpage.html  section 3.1
• Raoul Bott and Loring Tu. Differential Forms in Algebraic Topology, section 21.