Reider's theorem

From HandWiki
Revision as of 04:05, 26 December 2020 by imported>WikiGary (correction)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

In algebraic geometry, Reider's theorem gives conditions for a line bundle on a projective surface to be very ample.

Statement

Let D be a nef divisor on a smooth projective surface X. Denote by KX the canonical divisor of X.

  • If D2 > 4, then the linear system |KX+D| has no base points unless there exists a nonzero effective divisor E such that
    • [math]\displaystyle{ DE = 0, E^2 = -1 }[/math], or
    • [math]\displaystyle{ DE = 1, E^2 =0 }[/math];
  • If D2 > 8, then the linear system |KX+D| is very ample unless there exists a nonzero effective divisor E satisfying one of the following:
    • [math]\displaystyle{ DE = 0, E^2 = -1 }[/math] or [math]\displaystyle{ -2 }[/math];
    • [math]\displaystyle{ DE = 1, E^2 = 0 }[/math] or [math]\displaystyle{ -1 }[/math];
    • [math]\displaystyle{ DE = 2, E^2 = 0 }[/math];
    • [math]\displaystyle{ DE = 3, D = 3E, E^2 = 1 }[/math]

Applications

Reider's theorem implies the surface case of the Fujita conjecture. Let L be an ample line bundle on a smooth projective surface X. If m > 2, then for D=mL we have

  • D2 = m2 L2m2 > 4;
  • for any effective divisor E the ampleness of L implies D · E = m(L · E) ≥ m > 2.

Thus by the first part of Reider's theorem |KX+mL| is base-point-free. Similarly, for any m > 3 the linear system |KX+mL| is very ample.

References