Changing the base point in Abel-Jacobi divisor sums #
This file combines the point-level base-point-change API for the abstract Abel-Jacobi class with the divisor-level Abel-Jacobi sum.
For an order system whose principal divisors have weighted degree zero, and two weight-one
base points x₀ and y₀, the divisor-level Abel-Jacobi sums satisfy
AJ_{y₀}(D) = AJ_{x₀}(D) + deg(D) • ([x₀] - [y₀]).
Here deg(D) is the weighted degree for the chosen weight w. The unweighted specialization
is the same formula with the ordinary degree. This is the formal divisor-class bookkeeping
behind the later Abel maps D ↦ 𝒪_X(D - d·x₀) used in the Jacobian roadmap: changing the
normalizing base point translates the degree-d Abel map by d times the class of
[x₀] - [y₀].
This advances TauCetiRoadmap/JacobianChallenge/README.md, Layer A, "Pic⁰ X = ker deg (as
an abstract group)", and supplies a direct prerequisite for the Layer D/F Abel-map lane
D ↦ 𝒪_X(D - d·x₀). No external mathematics is vendored; the proofs reuse Tau Ceti's
existing weightedAbelJacobiDivisorClass, weightedBasepointChangeClass, and divisor-class
API.
Weighted base-point change for divisor sums #
Changing the base point in the weighted Abel-Jacobi sum adds the weighted degree times the base-point-change class.
Geometrically, for the residue-degree weight and rational base points x₀, y₀, this is the
formal divisor-class identity
[D - deg(D)y₀] = [D - deg(D)x₀] + deg(D)[x₀ - y₀] in Pic⁰.
In the class group, the difference between weighted Abel-Jacobi divisor sums with two
base points is the weighted degree times the class [x₀] - [y₀].
If a divisor has weighted degree zero, its weighted Abel-Jacobi sum is independent of the choice of weight-one base point.