Suppose we have an abelian category $\mathcal{A}$ and a topological space $X$. Let $\mathscr{F}^{\mathcal{A}}$ be the category of all sheaves of $\mathcal{A}$ objects over $X$. Is $\mathscr{F}^{\mathcal{A}}$ guaranteed to be an abelian category? I don't really have any evidence that this should be true other than that it happens to be true when $\mathcal{A}$ is Ab or $R$-Mod, and those are the principle abelian categories.

Of course, by the Freyd–Mitchell embedding theorem $\mathcal{A}$ is always equivalent to a full subcategory of $R$-Mod, but I'm not familiar enough with sheaves or the embedding theorem itself to know if the "subcategory" factor could play a problem here.

In the event that the answer to the posed question is "no", is there some extended criterion for $\mathscr{F}^{\mathcal{A}}$ to be abelian?


It seems that the answer to the posed question is "yes".

It is noted, but without proof (there might be no difficulty with a direct proof without using the Freyd-Mitchell embedding theorem or whatever non-trivial result), in Weibel An introduction to homological algebra (1.6 "More on abelian categories").

The category Sheaves(X) of sheaves on X with values in any abelian category is an abelian category, however it is not an abelian subcategory of Presheaves(X).

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.