Sheaves, Presheaves, and Sheaf Cohomology

One-line statement: A sheaf is a presheaf of local data that can be uniquely glued from compatible local pieces; sheaf cohomology measures the obstruction to passing from local to global solutions. Key result in words: The global sections functor

Γ (X, -)

is left exact but generally not exact; its right derived functors

H^{i} (X, F)

quantify the failure, and they reconcile the local-to-global principle that underlies modern geometry. Where it applies: Algebraic geometry, complex/differential geometry, algebraic topology, number theory, mathematical physics (gauge theory, string theory), and any setting requiring local-global comparison. Who/when: Jean Leray (1945, prison camp), refined by Henri Cartan, Jean-Pierre Serre, and Alexander Grothendieck (1950s–1960s).

A presheaf $F$ on a topological space $X$ assigns to each open set $U$ an abelian group (or set, ring, module) $F (U)$ together with restriction maps $ρ_{V}^{U} : F (U) \to F (V)$ for $V \subseteq U$ , functorially. A sheaf is a presheaf satisfying two gluing axioms: local sections that agree on overlaps glue uniquely to a global section. Sheaf cohomology arises because the category of sheaves of abelian groups has enough injectives, so the left-exact functor of global sections admits right derived functors. The surprising fact: cohomology computed via abstract injective resolutions, fine/flasque resolutions, Čech complexes, and (in the smooth case) the de Rham complex all coincide — a single invariant accessed through wildly different machinery, encoding obstructions ranging from the existence of meromorphic functions to the topology of $X$ itself.

The Core Insight

Local data + a consistency condition = a global object, up to a measurable obstruction. Presheaves record local data; the sheaf axioms enforce that this data behaves like functions (uniquely determined and freely gluable). Whenever you try to lift a local solution to a global one — a logarithm, a section of a line bundle, a primitive of a closed form — the failure is organized into a graded abelian group

H^{∙} (X, F)

. Cohomology is the bookkeeping device for everything that is locally trivial but globally non-trivial.

Formal Statement

Presheaf

Let $X$ be a topological space and $Open (X)$ the category whose objects are open sets and whose morphisms are inclusions. A presheaf of abelian groups is a contravariant functor

F : Open (X)^{op} ⟶ Ab .

⟳A contravariant functor reverses arrows; an inclusion

V ↪ U

yields the restriction

ρ_{V}^{U} : F (U) \to F (V)

Concretely $F$ assigns to each open $U$ an abelian group $F (U)$ (the sections over $U$ , written $Γ (U, F)$ ) and to each inclusion $V \subseteq U$ a homomorphism $ρ_{V}^{U}$ such that

ρ_{U}^{U} = id_{F (U)}, ρ_{W}^{V} \circ ρ_{V}^{U} = ρ_{W}^{U} (W \subseteq V \subseteq U) .

⟳These are precisely the functor identity and composition axioms transcribed for inclusions.

We write $s ∣_{V} := ρ_{V}^{U} (s)$ for $s \in F (U)$ .

Sheaf

A presheaf $F$ is a sheaf if for every open $U$ and every open cover $U = ⋃_{i \in I} U_{i}$ :

(Separation / Identity) If

s, t \in F (U)

satisfy

s ∣_{U_{i}} = t ∣_{U_{i}}

for all

i

, then

s = t

. (Gluing / Locality) Given

s_{i} \in F (U_{i})

with

s_{i} ∣_{U_{i} \cap U_{j}} = s_{j} ∣_{U_{i} \cap U_{j}}

for all

i, j

, there exists

s \in F (U)

with

s ∣_{U_{i}} = s_{i}

Equivalently, the following sequence is exact (an equalizer) for every open cover:

0 \to F (U) res i \prod F (U_{i}) \xrightrightarrows [β] α i, j \prod F (U_{i} \cap U_{j}),

where

α (s_{i})_{ij} = s_{i} ∣_{U_{ij}}

and

β (s_{i})_{ij} = s_{j} ∣_{U_{ij}}

, with

U_{ij} := U_{i} \cap U_{j}

⟳Injectivity of

res

encodes separation; equality

α = β

on the image plus surjectivity onto the equalizer encodes gluing.

Stalk and Sheafification

The stalk of $F$ at $x \in X$ is the filtered colimit over open neighborhoods $U ∋ x$ :

F_{x} := U ∋ x lim F (U),

whose elements are germs

[s]_{x}

. The sheafification

F^{+}

is the unique sheaf equipped with

θ : F \to F^{+}

inducing isomorphisms on all stalks and universal among morphisms from

F

to sheaves.

⟳Stalks are computed pointwise as germs; sheafification is the left adjoint to the inclusion

Sh (X) ↪ PSh (X)

, so it is determined by a universal property.

Existence of Sheaf Cohomology

The category $Sh (X, Ab)$ of sheaves of abelian groups is an abelian category with enough injectives. The global sections functor $Γ (X, -) : Sh (X, Ab) \to Ab$ is left exact. Its right derived functors

H^{i} (X, F) := R^{i} Γ (X, F) = H^{i} (Γ (X, I^{∙})),

computed from any injective resolution

0 \to F \to I^{0} \to I^{1} \to \dots

, are well-defined up to canonical isomorphism, with

H^{0} (X, F) = Γ (X, F)

⟳Enough injectives lets every

F

embed in an injective resolution; left exactness of

Γ

guarantees

H^{0} = Γ

; derived functor independence of resolution is the standard homological algebra comparison theorem.

Long Exact Sequence

A short exact sequence of sheaves $0 \to F^{'} \to F \to F^{''} \to 0$ induces a natural long exact sequence

0 \to H^{0} (X, F^{'}) \to H^{0} (X, F) \to H^{0} (X, F^{''}) δ H^{1} (X, F^{'}) \to \dots

\dots \to H^{i} (X, F^{'}) \to H^{i} (X, F) \to H^{i} (X, F^{''}) δ H^{i + 1} (X, F^{'}) \to \dots

⟳A short exact sequence of complexes (via injective resolutions and the horseshoe lemma) yields the snake-lemma connecting maps

δ

The connecting map $δ$ is precisely the local-to-global obstruction: an element of $H^{0} (X, F^{''})$ is a global section that lifts locally to $F$ ; $δ$ measures whether it lifts globally.

Why It Holds / Derivation

proof[Sheaf cohomology is well-defined and computable via acyclic resolutions] Step 1 — Abelian structure. Kernels and cokernels in $Sh (X)$ are computed stalkwise. The kernel presheaf $U \mapsto ker (F (U) \to G (U))$ is already a sheaf, but the cokernel presheaf need not be; one sheafifies it. Since sheafification is exact (it preserves stalks and stalkwise exactness), $Sh (X)$ is abelian.

Step 2 — Enough injectives. For each

x

, choose an injective abelian group

I_{x}

containing

F_{x}

. Form the Godement / skyscraper sheaf

I = x \in X \prod (i_{x})_{*} I_{x}, (i_{x})_{*} I_{x} (U) = {I_{x} 0 x \in U x \in / U .

⟳A product of injective skyscraper sheaves is injective because

Hom (G, (i_{x})_{*} I_{x}) ≅ Hom (G_{x}, I_{x})

, which is exact in

G

since

I_{x}

is injective and the stalk functor is exact.

The natural map $F ↪ I$ is injective on stalks, hence a monomorphism; iterating produces an injective resolution.

Step 3 — Left exactness of $Γ$ . Applying

Γ (X, -)

0 \to F^{'} \to F \to F^{''}

preserves kernels because a global section killed in

F

is killed in

F^{'}

stalkwise, hence is zero by the separation axiom. Right exactness fails: a global section of

F^{''}

need not lift globally. Step 4 — Acyclic resolutions suffice. A sheaf

A

is $Γ$ -acyclic if

H^{i} (X, A) = 0

for

i > 0

. Flasque (flabby), soft, and fine sheaves are acyclic. By the comparison theorem of homological algebra, any acyclic resolution

0 \to F \to A^{∙}

computes the same cohomology:

H^{i} (X, F) ≅ H^{i} (Γ (X, A^{∙})) .

⟳An acyclic resolution maps to an injective resolution by a chain map unique up to homotopy; applying

Γ

and using

H^{> 0} (A^{j}) = 0

shows the induced map on cohomology is an isomorphism via a dimension-shifting induction.

Step 5 — Flasque sheaves are acyclic. A flasque sheaf has surjective restriction maps

F (X) \to F (U)

. If

0 \to F \to G \to H \to 0

with

F

flasque, then

Γ

stays exact (sections lift because restrictions are surjective), and one shows by induction that

H^{> 0} = 0

■

Jean Leray | 1906–1998

Leray invented sheaves, the spectral sequence, and the notion of cohomology with supports while a prisoner of war (1940–1945) at Oflag XVII-A in Austria. To avoid being put to work on military applications of his fluid-dynamics expertise, he reinvented himself as a topologist. His lectures, smuggled out and published in 1945–1946, introduced faisceaux (sheaves) as a tool to track how local cohomological data varies. Henri Cartan's Paris seminar (1948–1951) reorganized the theory into its modern stalk-and-section form; Serre's FAC (1955) and Grothendieck's Tôhoku paper (1957) recast everything in the language of derived functors and abelian categories.

Building Intuition

Think of a presheaf as data attached to regions, and a sheaf as data that behaves like functions: you can chop a function up over a cover and reassemble it uniquely. The classic failure of gluing — and hence the birth of cohomology — is the logarithm. On any small disk in $C^{\times}$ you can define $lo g z$ ; these local logs agree on overlaps up to a constant $2 π ik$ , but no global continuous $lo g$ exists on the punctured plane. That obstruction is a class in $H^{1}$ .

The hierarchy of cohomology degrees has a memorable reading:

$H^{0}$ : global solutions that actually exist (e.g., global sections, locally constant functions).
$H^{1}$ : first obstruction to gluing — torsors, line bundles, extensions, the failure of "locally solvable $\Rightarrow$ globally solvable."
$H^{\geq 2}$ : higher coherence obstructions — gerbes, deformation obstructions, secondary glueing data.

A sheaf is fine or soft when it has partitions of unity (true for smooth functions): then any local data can be averaged into global data, and all higher cohomology vanishes. Cohomology is thus a measure of the rigidity of the sheaf against being smoothed away.

Local sections over a cover failing to glue: the genesis of $H^1$. Local logarithms (gold) agree on overlaps up to a period, but no global branch exists on the annulus.

Worked Examples

Čech cohomology of the circle $S^1$ with the constant sheaf $\underline{\mathbb{Z}}$

Given:

X = S^{1}

, sheaf

\underline{Z}

of locally constant integer-valued functions. Cover

S^{1}

by two connected arcs

U_{0}, U_{1}

whose intersection

U_{0} \cap U_{1}

is two disjoint arcs

A ⊔ B

. Step 1 — Sections.

\underline{Z} (U_{0}) = Z

\underline{Z} (U_{1}) = Z

(each arc connected), and

\underline{Z} (U_{0} \cap U_{1}) = Z \oplus Z

(two components). Step 2 — Čech complex.

0 \to Z \oplus Z d^{0} Z \oplus Z \to 0, d^{0} (a, b) = (a - b, a - b) .

Step 3 — $H^{0}$ .

ker d^{0} = {(a, b) : a = b} ≅ Z

. ✓ (Connected space: one global constant.) Step 4 — $H^{1}$ .

coker d^{0} = (Z \oplus Z) / {(t, t)} ≅ Z

via

(x, y) \mapsto x - y

. Answer:

H^{0} (S^{1}, \underline{Z}) = Z, H^{1} (S^{1}, \underline{Z}) = Z, H^{\geq 2} = 0

matching singular cohomology

H^{∙} (S^{1})

. The generator of

H^{1}

is the "winding" cocycle.

Cohomology of $\mathcal{O}(d)$ on $\mathbb{P}^1$ over a field $k$

Given: Line bundles

O (d)

P_{k}^{1} = U_{0} \cup U_{1}

with affine charts

U_{0} = Spec k [t]

U_{1} = Spec k [t^{- 1}]

t = x_{1} / x_{0}

. Step 1 — Sections. Global sections of

O (d)

are degree-

d

homogeneous polynomials in

x_{0}, x_{1}

H^{0} (P^{1}, O (d)) = {k^{d + 1} 0 d \geq 0 d < 0.

Step 2 — Čech on two charts. With

O (d) (U_{0}) = k [t]

O (d) (U_{1}) = k [t^{- 1}]

O (d) (U_{01}) = k [t, t^{- 1}]

H^{1} = \frac{k [ t , t ^{- 1} ]}{t ^{d} k [ t ] + k [ t ^{- 1} ]} (after twisting by t^{d}) .

Step 3 — Count. The quotient is spanned by Laurent monomials

t^{- 1}, t^{- 2}, \dots, t^{d + 1}

, giving dimension

- d - 1

when

d \leq - 2

. Answer:

dim H^{0} = max (d + 1, 0), dim H^{1} = max (- d - 1, 0)

This satisfies Serre duality

H^{1} (O (d)) ≅ H^{0} (O (- 2 - d))^{\lor}

(note

K_{P^{1}} = O (- 2)

), and Euler characteristic

χ (O (d)) = d + 1

, consistent with Riemann–Roch on genus

0

Common Misconceptions

A presheaf cokernel is not a sheaf. If

ϕ : F \to G

is a sheaf morphism, the presheaf

U \mapsto coker (F (U) \to G (U))

generally violates gluing; the sheaf cokernel is its sheafification. Consequently,

ϕ

can be surjective as a sheaf map (surjective on all stalks) while

Γ (X, ϕ)

is not surjective — this is exactly what

H^{1}

detects. Sheaf cohomology is not always the topology of $X$ .

H^{i} (X, F)

depends on both

X

and

F

. The constant sheaf recovers singular cohomology only for sufficiently nice (e.g. locally contractible) spaces; coherent sheaves on a scheme see algebraic, not topological, structure. Čech $\neq =$ derived functor cohomology in general. They agree on paracompact spaces (Leray's theorem with acyclic covers) and for quasi-coherent sheaves on separated schemes (affine covers), but Čech cohomology can differ on pathological spaces. Always check the cover is acyclic (a Leray cover) before trusting Čech computations.

Constant presheaf is not a sheaf

Let $X$ be a space with two points (or any disconnected open set) and define the constant presheaf $F (U) = Z$ for all nonempty $U$ , with identity restriction maps. Take $U = U_{1} ⊔ U_{2}$ disconnected. Sections $1 \in F (U_{1})$ and $2 \in F (U_{2})$ trivially agree on the empty overlap, but there is no element of $F (U) = Z$ restricting to both. Gluing fails. The sheafification is the locally constant sheaf $\underline{Z}$ with $\underline{Z} (U) = Z^{# {components of U}}$ . This single example explains why $H^{0}$ counts connected components.

Connections

Sheaf cohomology is the keystone linking many theories:

De Rham Cohomology: The Poincaré lemma makes the de Rham complex $0 \to \underline{R} \to Ω^{0} \to Ω^{1} \to \dots$ a fine (hence acyclic) resolution of the constant sheaf, so $H_{dR}^{i} (M) ≅ H^{i} (M, \underline{R})$ — the de Rham theorem.
Dolbeault Cohomology: For a complex manifold, the $\overset{ˉ}{\partial}$ -Poincaré lemma gives $H^{q} (X, Ω^{p}) ≅ H_{\overset{ˉ}{\partial}}^{p, q} (X)$ .
Čech Cohomology: A combinatorial model agreeing with derived cohomology on Leray covers; the bridge is the Čech-to-derived spectral sequence.
Singular Cohomology: Equals $H^{∙} (X, \underline{Z})$ on locally contractible spaces.
Grothendieck Topology: Replaces open covers by abstract covering families, yielding sheaves on a site and étale cohomology.
Spectral Sequences: The Leray spectral sequence $H^{p} (Y, R^{q} f_{*} F) \Rightarrow H^{p + q} (X, F)$ computes cohomology through a continuous map $f : X \to Y$ .
Derived Categories: The natural home of $R Γ$ , where sheaf cohomology becomes hyper-derived functor cohomology.

Real-World Applications

Riemann–Roch & algebraic geometry: Dimensions $h^{0} (X, O (D))$ of linear systems of divisors — controlling embeddings of curves and surfaces — are computed via $χ (O (D))$ and the vanishing $H^{1} = 0$ (Kodaira vanishing).
Hodge theory & period domains: The Hodge decomposition $H^{n} (X, C) = ⨁ H^{q} (X, Ω^{p})$ underlies the classification of complex projective varieties and mirror symmetry.
Gauge theory / instantons (physics): Principal $G$ -bundles are classified by $H^{1} (X, \underline{G})$ ; characteristic classes (Chern, Pontryagin) live in $H^{∙} (X, Z)$ and detect topological charge.
String theory & D-branes: B-fields are classes in $H^{3} (X, Z)$ (gerbes, $H^{2}$ with $O^{\times}$ coefficients); D-brane charges live in twisted K-theory derived from sheaf-theoretic data.
Number theory (étale cohomology): Grothendieck's site-theoretic sheaf cohomology produces $ℓ$ -adic cohomology, proving the Weil conjectures (Deligne) and underpinning the Langlands program.
Deformation theory: $H^{1} (X, T_{X})$ parametrizes infinitesimal deformations of a variety; $H^{2}$ holds the obstructions — used in moduli space construction.
Topological data analysis & sensor networks: Cellular sheaves model distributed data; $H^{0}$ detects global consensus, $H^{1}$ detects inconsistency/coverage holes in robotic sensor coverage problems.

Historical Development

1945–1946 — Jean Leray, recently freed from an Oflag POW camp, publishes the first definition of faisceaux and spectral sequences in the Comptes Rendus and his Collège de France course.
1948–1951 — Henri Cartan's Paris seminar (Séminaire Cartan) systematizes sheaves via stalks and sections; the étale-space definition and the gluing axioms take modern form.
1953 — Cartan & Eilenberg's Homological Algebra provides the derived-functor language soon applied to sheaves.
1955 — Jean-Pierre Serre's FAC (Faisceaux Algébriques Cohérents) introduces coherent sheaves and their cohomology into algebraic geometry, proving finiteness and vanishing theorems.
1956 — Hirzebruch uses sheaf cohomology to prove the Riemann–Roch–Hirzebruch theorem.
1957 — Alexander Grothendieck's Tôhoku paper recasts sheaf cohomology as derived functors in abelian categories (axiom AB5, enough injectives), removing all reliance on covers.
1958–1964 — Grothendieck and the SGA seminars introduce Grothendieck topologies, sites, topoi, and étale cohomology, aiming at the Weil conjectures.
1973 — Pierre Deligne completes the proof of the Weil conjectures using $ℓ$ -adic étale cohomology.
1990s–present — Derived/triangulated categories, perverse sheaves (BBD), and persistent sheaf cohomology in applied topology extend the framework.

Subtleties & Limits

Hodge conjecture (Clay Millennium Problem). On a smooth projective complex variety, are all rational

(p, p)

-Hodge classes in

H^{2 p} (X, Q)

algebraic (i.e., combinations of cohomology classes of subvarieties)? This is fundamentally a question about which sheaf-cohomological classes arise geometrically. Still open in general. Standard conjectures (Grothendieck). A web of conjectures about algebraic cycles and motives that would unify the various Weil cohomology theories (de Rham,

ℓ

-adic, crystalline) into a universal motivic cohomology; largely open. Computational limits. While

H^{i}

of coherent sheaves on projective space is algorithmically computable (Gröbner bases, Macaulay2), cohomology of arbitrary sheaves on general spaces is uncomputable; even for nice cases, vanishing theorems beyond Kodaira (e.g. effective bounds, characteristic-

p

failures of Kodaira vanishing à la Raynaud) remain delicate. Non-abelian cohomology.

H^{1} (X, \underline{G})

for non-abelian

G

classifies torsors but is only a pointed set; a satisfactory general

H^{\geq 2}

(higher gerbes,

\infty

-stacks) is an active area where the "right" definitions are still being negotiated in derived/higher categorical frameworks.

Sheaves, Presheaves, and Sheaf Cohomology

The Core Insight

Formal Statement

Presheaf

Sheaf

Stalk and Sheafification

Existence of Sheaf Cohomology

Long Exact Sequence

Why It Holds / Derivation

Jean Leray | 1906–1998

Building Intuition

Local sections over a cover failing to glue: the genesis of $H^1$. Local logarithms (gold) agree on overlaps up to a period, but no global branch exists on the annulus.

Worked Examples

Čech cohomology of the circle $S^1$ with the constant sheaf $\underline{\mathbb{Z}}$

Cohomology of $\mathcal{O}(d)$ on $\mathbb{P}^1$ over a field $k$

Common Misconceptions

Constant presheaf is not a sheaf

Connections

Real-World Applications

Historical Development

Subtleties & Limits

See Also

Before this

Where this leads