Cohomologically rigid vector fields: the Katok conjecture in dimension 3 (Q2272064): Difference between revisions

On a closed orientable manifold \(M\) of dimension \(d\), a smooth vector field \(X\) is called ``cohomologically rigid'' when, for any \(\xi\in C^\infty(M,\mathbb{R})\), there exists \(u\in C^\infty(M,\mathbb{R})\) and \(c\in\mathbb{R}\) such that \({\mathcal L}_X u=\xi-c\), where \({\mathcal L}_X\) is the Lie derivative in the \(X\) direction. The \textit{A. Katok} conjecture [Proc. Int. Congr. Math., Warszawa 1983, Vol. 2, 1245--1253 (1984; Zbl 0563.58016)] is that every cohomologically rigid vector field is smoothly conjugate to a Diophantine vector field on the torus \(\mathbb{T}^d\). A Diophantine vector field consists of vectors \(\alpha= (\alpha_1,\alpha_2, \dots,\alpha_d)\) for which there exist positive constants \(C\) and \(\tau\) such that \[ \Biggl|\sum^d_{i=1} \alpha_i p_i\Biggr|> C\Biggl(\max_{1\leq i\leq d}|p_i|\Biggr)^{-\tau} \] for every \(p= (p_1,p_2,\dots, p_d)\in \mathbb{Z}^d-\{0\}\). The author's main result is a proof that Katok's conjecture is true in three dimensions. The background for this work is as follows: If \(G\) is a Lie group acting on a manifold via a smooth action \(\Phi: G\times M\to M\), then questions about the dynamics of \(\Phi\) can often be answered by looking at the first cohomology group \(H^1(\Phi)\). Commonly \(H^1(\Phi)\) is infinite-dimensional. When \(H^1(\Phi)\) is ``small'' -- its dimension finite -- it is usually associated with some kind of rigidity of the group action \(\Phi\).