← Back to arXiv
arXivAlgebraic GeometryarXiv:2609.25172

Local Logarithmic Cartier Transform

The Cartier transform is a powerful tool in algebraic geometry that creates a correspondence between two different kinds of geometric objects defined over fields with positive characteristic (like fields of integers modulo a prime number). On one side are objects called modules with Higgs fields, and on the other side are modules with flat connections. These two types of objects encode geometric and differential information in different ways, and having a bridge between them is very useful for understanding the structure of algebraic varieties. The original version of this transform was developed by Ogus and Vologodsky, and a local version was later worked out by Shiho.

The paper extends Shiho's local version of the Cartier transform to a more general setting called the logarithmic setting. Logarithmic geometry is a framework that allows mathematicians to work with spaces that have certain kinds of boundary behavior or singularities in a controlled way, essentially by keeping track of "where things go to infinity" in a structured manner. The authors show that under reasonable technical conditions, including the ability to lift a certain map called the Frobenius to a richer algebraic setting, one can construct a well-behaved functor that translates between the two categories of geometric objects in this logarithmic context. The functor is fully faithful, meaning it preserves all the relevant structure without any loss of information.

A key ingredient in the paper is a new descent theorem for morphisms in logarithmic flat geometry, building on earlier foundational work by Kato. Descent theorems are results that let you assemble local information into global geometric objects, and proving one in this setting required substantial technical work. This paper is described as the first in a series of three, with the broader goal being a complete generalization of the Ogus-Vologodsky Cartier transform to logarithmic geometry, which would open up new tools for studying a wide range of algebraic varieties with boundary structure.

Read original →