Let $X^n$ be a Stein manifold (complex submanifold in $\mathbb{C}^N$ for some large $N$). Let $D = \{(z,z)\in X\times X: z\in X\}$ be the diagonal in $X\times X$. I'm looking for some holomorphic vector bundle $B$ of dimension $n$ over $X\times X$ and a global section $s : X\times X \to B$ such that $D = s^{-1}(0)$.

There's a construction by Henkin-Leiterer as follows: Take $B = \pi^* TX $ where $\pi: X \times X \to X$ is the projection onto its first variable and $\pi^*$ is the pull-back. Let $F : X\to \mathbb{C}^N$ be an embedding for Stein manifold $X$. Then $d F : T(X)\to X\times \mathbb{C}^N$ is an injection between holomorphic vector bundles. Use Cartan B, one can show that there's a holomorphic left inverse $(dF)^{-1}: X\times \mathbb{C}^N \to T(X)$ such that $(dF)^{-1} (dF)= Id$. Let $(z,\zeta)\in X \times X$, define $s(z,\zeta)= (\pi)^*(d F)^{-1} (z, F(z)- F(\zeta))$. One sees that $D\subset s^{-1}(0)$. Henkin and Leiterer showed that the extra zero locus (if exists) is disjoint from D by proving that for fixed $z\in X$, $s(z,\zeta)$ is locally biholomorphism in the second variable.

My question is, can one show that there's no extra zero locus? If not, does there exist such pairing $(B,s)$ such that $s^{-1}(0)=D$ without extra zero locus? Thanks!


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.