Inverse image functor
In mathematics, specifically in algebraic topology and algebraic geometry, an inverse image functor is a contravariant construction of sheaves; here “contravariant” in the sense given a map [math]\displaystyle{ f : X \to Y }[/math], the inverse image functor is a functor from the category of sheaves on Y to the category of sheaves on X. The direct image functor is the primary operation on sheaves, with the simplest definition. The inverse image exhibits some relatively subtle features.
Image functors for sheaves |
---|
direct image f∗ |
inverse image f∗ |
direct image with compact support f! |
exceptional inverse image Rf! |
[math]\displaystyle{ f^* \leftrightarrows f_* }[/math] |
[math]\displaystyle{ (R)f_! \leftrightarrows (R)f^! }[/math] |
Base change theorems |
Definition
Suppose we are given a sheaf [math]\displaystyle{ \mathcal{G} }[/math] on [math]\displaystyle{ Y }[/math] and that we want to transport [math]\displaystyle{ \mathcal{G} }[/math] to [math]\displaystyle{ X }[/math] using a continuous map [math]\displaystyle{ f\colon X\to Y }[/math].
We will call the result the inverse image or pullback sheaf [math]\displaystyle{ f^{-1}\mathcal{G} }[/math]. If we try to imitate the direct image by setting
- [math]\displaystyle{ f^{-1}\mathcal{G}(U) = \mathcal{G}(f(U)) }[/math]
for each open set [math]\displaystyle{ U }[/math] of [math]\displaystyle{ X }[/math], we immediately run into a problem: [math]\displaystyle{ f(U) }[/math] is not necessarily open. The best we could do is to approximate it by open sets, and even then we will get a presheaf and not a sheaf. Consequently, we define [math]\displaystyle{ f^{-1}\mathcal{G} }[/math] to be the sheaf associated to the presheaf:
- [math]\displaystyle{ U \mapsto \varinjlim_{V\supseteq f(U)}\mathcal{G}(V). }[/math]
(Here [math]\displaystyle{ U }[/math] is an open subset of [math]\displaystyle{ X }[/math] and the colimit runs over all open subsets [math]\displaystyle{ V }[/math] of [math]\displaystyle{ Y }[/math] containing [math]\displaystyle{ f(U) }[/math].)
For example, if [math]\displaystyle{ f }[/math] is just the inclusion of a point [math]\displaystyle{ y }[/math] of [math]\displaystyle{ Y }[/math], then [math]\displaystyle{ f^{-1}(\mathcal{F}) }[/math] is just the stalk of [math]\displaystyle{ \mathcal{F} }[/math] at this point.
The restriction maps, as well as the functoriality of the inverse image follows from the universal property of direct limits.
When dealing with morphisms [math]\displaystyle{ f\colon X\to Y }[/math] of locally ringed spaces, for example schemes in algebraic geometry, one often works with sheaves of [math]\displaystyle{ \mathcal{O} Y }[/math]-modules, where [math]\displaystyle{ \mathcal{O}_Y }[/math] is the structure sheaf of [math]\displaystyle{ Y }[/math]. Then the functor [math]\displaystyle{ f^{-1} }[/math] is inappropriate, because in general it does not even give sheaves of [math]\displaystyle{ \mathcal{O}_X }[/math]-modules. In order to remedy this, one defines in this situation for a sheaf of [math]\displaystyle{ \mathcal O_Y }[/math]-modules [math]\displaystyle{ \mathcal G }[/math] its inverse image by
- [math]\displaystyle{ f^*\mathcal G := f^{-1}\mathcal{G} \otimes_{f^{-1}\mathcal{O}_Y} \mathcal{O}_X }[/math].
Properties
- While [math]\displaystyle{ f^{-1} }[/math] is more complicated to define than [math]\displaystyle{ f_{\ast} }[/math], the stalks are easier to compute: given a point [math]\displaystyle{ x \in X }[/math], one has [math]\displaystyle{ (f^{-1}\mathcal{G})_x \cong \mathcal{G}_{f(x)} }[/math].
- [math]\displaystyle{ f^{-1} }[/math] is an exact functor, as can be seen by the above calculation of the stalks.
- [math]\displaystyle{ f^* }[/math] is (in general) only right exact. If [math]\displaystyle{ f^* }[/math] is exact, f is called flat.
- [math]\displaystyle{ f^{-1} }[/math] is the left adjoint of the direct image functor [math]\displaystyle{ f_{\ast} }[/math]. This implies that there are natural unit and counit morphisms [math]\displaystyle{ \mathcal{G} \rightarrow f_*f^{-1}\mathcal{G} }[/math] and [math]\displaystyle{ f^{-1}f_*\mathcal{F} \rightarrow \mathcal{F} }[/math]. These morphisms yield a natural adjunction correspondence:
- [math]\displaystyle{ \mathrm{Hom}_{\mathbf {Sh}(X)}(f^{-1} \mathcal G, \mathcal F ) = \mathrm{Hom}_{\mathbf {Sh}(Y)}(\mathcal G, f_*\mathcal F) }[/math].
However, the morphisms [math]\displaystyle{ \mathcal{G} \rightarrow f_*f^{-1}\mathcal{G} }[/math] and [math]\displaystyle{ f^{-1}f_*\mathcal{F} \rightarrow \mathcal{F} }[/math] are almost never isomorphisms. For example, if [math]\displaystyle{ i\colon Z \to Y }[/math] denotes the inclusion of a closed subset, the stalk of [math]\displaystyle{ i_* i^{-1} \mathcal G }[/math] at a point [math]\displaystyle{ y \in Y }[/math] is canonically isomorphic to [math]\displaystyle{ \mathcal G_y }[/math] if [math]\displaystyle{ y }[/math] is in [math]\displaystyle{ Z }[/math] and [math]\displaystyle{ 0 }[/math] otherwise. A similar adjunction holds for the case of sheaves of modules, replacing [math]\displaystyle{ i^{-1} }[/math] by [math]\displaystyle{ i^* }[/math].
References
- Iversen, Birger (1986), Cohomology of sheaves, Universitext, Berlin, New York: Springer-Verlag, ISBN 978-3-540-16389-3. See section II.4.
![]() | Original source: https://en.wikipedia.org/wiki/Inverse image functor.
Read more |