Search the corpus

    Ampleness of a divisor on a curve

    PR-DIVAMPCURVE

    On a smooth projective curve, positive degree is ample, degree $2g$ kills base points, and degree $2g+1$ is very ample.Suggest a slogan

    Let \(D\) be a divisor on a smooth projective curve \(X\) of genus \(g\) over \(k = \bar{k}\).

    • \(D\) is ample if and only if \(\deg D > 0\).

    • \(\deg D \geq 2g\) implies \(\abs{D}\) is base-point free.

    • \(\deg D \geq 2g+1\) implies \(D\) is very ample.

    On a curve ampleness collapses to a single integer, and this is the cleanest place to see the ample-against-very-ample distinction: positivity of the degree is ample, but very ampleness needs the degree to clear a genus-dependent threshold. Raising the degree past \(2g+1\) is exactly what taking powers achieves, which is the definition of ample in this case.

    The thresholds are what Riemann–Roch gives: once \(\deg D > 2g-2\) the correction term vanishes and \(h^0\) is computable, so the conditions for base points and for separation become arithmetic. Both bounds are sharp on an elliptic curve, where \(g = 1\) and degree \(2\) gives the base-point-free double cover of \(\PP^1\) while degree \(3\) first embeds.

    When is a divisor on a curve ample?
    What degree makes a divisor on a curve base-point free? Very ample?

    Dependencies

    • D-DIVAMPLEAmple and very ample divisorsuses
    • PR-Y5S7VThe degree of a divisor on a projective curveuses

    What links to this