The Weil bound and arithmetic positivity
To read this chapter you need Arakelov geometry: Hermitian line bundles, arithmetic intersection numbers and volumes, and the slopes of Hermitian lattices, with a little about CM fields and Shimura varieties.
The Weil bound chapter is not about Weil's explicit formula. It follows the branch deep_slim_use, which proves a positivity statement of Arakelov geometry from eight published theorems. Its conclusion DeepStR (defined below) bounds from below a mean slope of small cusp sections on unitary Shimura varieties. It also lists the timeline steps with no counterpart in the code.
Read first
- Number fields and their integers: the discriminant, and the root discriminant of the error term.
- Kinds of number fields. is a CM field, abelian over in the volume theorem.
- The logarithmic derivative and quadratic symbols give for the quadratic character .
- That term comes from the Chowla-Selberg formula, in Step 4.
- Abelian varieties and abelian schemes. SlopeBoundSt is about the Lie algebra of a semi-abelian scheme.
- Dieudonné modules and the Hodge-Tate sequence: one input is an exactness result of Fargues.
- Schemes and varieties. The arithmetic variety sits in a product of projective spaces.
- Vanishing theorems and resolution states three inputs: resolution of singularities, principalisation and asymptotic Riemann-Roch for nef divisors.
- Shimura varieties and toroidal compactifications, for the tower of integral models.
- Hermitian line bundles: semipositive metrics and the Poincaré-Lelong formula.
- Arithmetic varieties and intersection numbers: the arithmetic intersection number , compared with counts of sections.
- Heights gives Zhang's successive minima.
- Arithmetic volume and bigness covers Moriwaki and Yuan, the arithmetic Hilbert-Samuel theorem and the arithmetic Kleiman criterion.
- Hermitian lattices and slopes: the maximal slope, the absolute Siegel lemma and, for the Notes, the Bost slope inequality.
- The Hodge bundle and Zarhin's trick, the deepest, for the Hodge metric and Zarhin's input.
The new objects
- DeepStR says that MeanAll holds with for each tower object.
- MeanAll says: for each weight with a gap, the volume slopes of the components have mean at least . A volume slope grows with the number of lattice sections of small sup norm, so there are many small cusp sections.
- MoriwakiBigStatement is the inequality of Moriwaki and Yuan for a nef Hermitian line bundle. HSUse is the upper half of arithmetic Hilbert-Samuel with the arithmetic Kleiman criterion. Both are proved in Lean, for intersection numbers given by axioms.
- An absolute Siegel lemma gives each Hermitian lattice a vector of large degree over some finite extension. SlopeBoundSt bounds of the Lie algebra of a semi-abelian scheme by .
- GoodTowerExistsX gives a tower of integral models of a unitary Shimura variety. VolumeTheoremX bounds from below the mean intersection slope of the Hodge line bundles of weight on it.
The path through the chapter
- Step 2 gets small sections from intersection numbers by Moriwaki and Yuan, and the reverse bound by Hilbert-Samuel and Zhang's successive minima.
- Step 3 gets the absolute Siegel lemma, and from it SlopeBoundSt, in place of a Bost slope inequality.
- Step 4 computes intersection numbers on the tower. The Chowla-Selberg formula gives the term .
- Step 5 turns the intersection slope into a volume slope, again by Moriwaki and Yuan, and so proves DeepStR.
The source of this page. Markdown. Something is wrong? Report it.