Generally, if we have any space $X$, then multiplication by any odd class $\eta$ makes $H^*(X)$ a complex (usually called Aomoto complex) $A_{\eta}$ because cup product is graded commutative. It is pretty interesting structure: for example, for space $M$ with 1-formal $\pi_1$ tangent cone at 0 of $V \subset H^1(M, \mathbb C)$, $V := \{\alpha ; \,H^1(A_{\alpha}) \neq 0 \}$ coincides with tangent cone at 1 of $R \subset Hom(\pi_1, \mathbb C^{\times})$, where $R := \{ \chi \in Hom(\pi_1, \mathbb C^{\times}); \, \mathrm{dim} \, H^1(M, \mathbb C_{\chi}) \neq 0 \}$. Also, Betti numbers of $p$-torsion local systems are bounded by Aomoto Betti numbers with $\mathbb Z/p$ coefficients.

But if we have a distinguished closed $2k+1$-form $\eta$ on a manifold, then we have one special complex $A_{\eta}$. There are a few examples: locally conformally Kähler/symplectic manifolds with Lee form, Lie groups with their Chern-Simons 3-form, Cartan 1-form, canonical 3-cocycle of ss group, deforming 3-form for generalized complex manifold and so on.

So the quesion is: are there nice geometric interpretations of $H^*(A_{\theta})$ or $H^*(A_{d\omega})$ for LCK manifold $M$? Are they isomorphic to cohomology of particular sheaf on $M$? Which conditions imply or implied by their vanishing in some degree? (I'm mainly interested in LCK/LCS case, but happy to see aswers for other geometric structures.)

Also, as $(H^*(M), \wedge \eta)$ is in some sense linearization of twisted de Rham $(\Omega^*(M), d + \wedge \eta)$, we may ask whether twisted cohomology can be recovered via some kind of spectral sequence from Aomoto complex (and if so, we can adjust some known things about Morse-Novikov cohomology to initial data).


Your Answer

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

Browse other questions tagged or ask your own question.