MathOverflow is a question and answer site for professional mathematicians. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

Let $f:X\to Y$ be a local complete intersection morphism (of schemes or varieties) of (relative) dimension $c$ everywhere. Is it true that $f^!\cong f^*[2c]$ (as a functor between the derived categories of etale constructible sheaves?), or does one really need smoothness here? What additional assumptions (or corrections) are needed here, and what is the best reference (for local complete intersection morphisms)?

Upd. As Torsten Ekedahl states, this is wrong. Still, are their any special properties of cohomology of local complete intersection morphisms (that are common with smooth morphisms, and distinguish them from general finite type morphisms)?

share|cite|improve this question
This is false already for $Y$ a point and $X$ not a homology manifold. – Torsten Ekedahl Sep 22 '11 at 7:18
Thank you! Then could you tell me: do local complete intersection morphism have any 'special cohomological' properties (similar to those of smooth morphisms)? – Mikhail Bondarko Sep 22 '11 at 8:14
For a variety $X$ with l.c.i. singularities the constant sheaf $\mathbb{Q}_{l,X}$ (with the appropriate shift) is perverse but I don't know if there is something one can say for arbitrary l.c.i. morphisms. – ulrich Sep 22 '11 at 8:57
Thanks!! Possibly this is all I need. – Mikhail Bondarko Sep 22 '11 at 9:36
This is obviously false in a very different situation than the one given by Torsten Ekedahl : Y is the affine line, X is the origin, f=i is the inclusion (a regular immersion), K is any complex on X-Y, j is the inclusion of X-Y in X, then i^!j_*K is always 0 but that's not true for i^*j_*K. On a less trivial note, if what you wanted were true just for the constant sheaf on Y, then it would mean that absolute purity holds for any regular closed immersion, not just for one between regular schemes. I'm a bit skeptical (though I can't give a counterexample right now). – Alex Sep 22 '11 at 22:56

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.