Skip to content

Sheaves

by gpt-5.2-codex
Learning objectives
  • Sheaves

Entry conditions

Use sheaves only when you can:

  • Start from a site (C,J)(C,J).
  • Define a presheaf on CC.
  • State a concrete gluing requirement for your data.

Definitions

  • A sheaf is a presheaf that satisfies the sheaf condition: compatible local data can be uniquely glued into global data.

Vocabulary (plain language)

  • Compatible: local pieces agree on overlaps.
  • Glue: combine compatible local pieces into one global piece.
  • Sheaf condition: the formal rule that guarantees unique gluing.

Symbols used

  • FF: a presheaf or sheaf.
  • {UiU}\{U_i \to U\}: a cover of UU.

Intuition

Sheaves formalize “local-to-global.” If you know data on all pieces of a cover and those pieces agree where they overlap, then there is one global piece of data that explains all the local pieces.

Worked example

Let CC be the open sets of a space and F(U)F(U) be continuous functions on UU. If functions agree on overlaps, there is a unique function on the union that restricts to each piece. This presheaf is a sheaf.

How to recognize the structure

  • You can define what it means for local data to agree.
  • You can define what it means to glue.
  • Gluing is unique when it exists.

Common mistakes

  • Using sheaves without a clear notion of overlap.
  • Assuming gluing is unique when it is not.

Minimal data

Relations

Authors
Date created
Teaches

Cite

@misc{gpt-5.2-codex2025-sheaves,
  author    = {gpt-5.2-codex},
  title     = {Sheaves},
  year      = {2025},
  url       = {https://emsenn.net/library/math/domains/topology/texts/sheaves/},
  publisher = {emsenn.net},
  license   = {CC BY-SA 4.0}
}