Is there a notion (for schemes or just locally ringed spaces) of cohomology with compact support? I guess there is for algebraic schemes over $\mathbf{C}$, but what about schemes in general? Does anybody have a good reference?

$\begingroup$ If you really mean the word "coherent" in the title, then the answer is no. E.g., you would want a $j_!$ morphism for $j$ an open embedding, but one can see that there's no left adjoint to $j^*$ in this setting (since it would imply tensor products commute with infinite products). $\endgroup$ – Moosbrugger Oct 13 '11 at 1:54
Yes there is. Take $S$ a scheme and take $f \colon X \to S$ a compactifiable morphism of schemes. By definition this means that there exists a proper $S$scheme Y which contains $X$ as an open subscheme. Then, given a compactification and a sheaf on $X$, you may define the cohomology with proper support of this sheaf as the cohomology of the pushforward of your sheaf to $Y$.
I think this works for many different cohomologies, but you need to check that the compact support cohomology does not depend on the chosen compactification. At least for etale sheaves I know this is so.
SGA 2 is the original reference. You can find it here
If X is a smooth scheme over complex numbers then you can consider $X_{an}$ as an complex analytic manifold and compute singular/ deRham/ simplicial cohomology with compact supports (this will be different from usual cohomology if X is not proper) On the Algebraic side there is etale cohomology with compact supports (which is defined by embedding X into a proper scheme...).
Comparison theorems tell you that etale cohomology with torsion coefficients agree with singular cohomology (with torsion coeff).
Any reference on etale cohomology will discuss this.
Ref:SGA 4.5, Milne: Etale Cohomology.

$\begingroup$ For a discussion of "Sheaf cohomology wih compact supports over any topological space" you may want to look up "Sheaf theory  Breadon" and " Sheaves in topologyA.Dimca". It should be noted that: Etale cohomology with compact support requires the existence of proper embedding (one shows that result is independent of a chosen embedding), so it is not a straightforward generalization of "Sheaf cohomology wih compact supports" to the etale site. $\endgroup$ – isildur Oct 13 '11 at 0:22