The transport Oka-Grauert principle for simple surfaces (Q6158342): Difference between revisions

On a Riemannian surface \((M,g)\) the unit tangent bundle \(SM\) is a \(S^1\)-bundle which has a natural frame of vector fields \(X,V,X_{\perp}=[X,V]\) where \(V\) is the induced field of the \(S^1\)-action and \(X\) is the geodesic vector field generated by the geodesic flow on \(M\). The paper under review considers a transport equation on \(SM\) determined by \(X\) and a Lie algebra-valued function \(\mathbb{A}\) called attenuation. When \(M\) has boundary the solution of this equation defines a nonabelian X-ray transform with source \(\mathbb{A}\) on the boundary of \(SM\). The injectivity, up to a gauge transformation, of the nonabelian X-ray transform for \(\mathbb{A}\) with values in \(u(n)\) and \(gl(n, \mathbb{C})\) is proven in [\textit{G. P. Paternain} et al., Geom. Funct. Anal. 22, No. 5, 1460--1489 (2012; Zbl 1256.53021); \textit{G. P. Paternain} and \textit{M. Salo}, ``The non-Abelian X-ray transform on surfaces'', Preprint, \url{arXiv:2006.02257}]. One of the results is a description of the range of the X-ray transform on a simple Riemannian surface with boundary when \(\mathbb{A}\) is \(u(n)\)-valued. Here the surface \((M, g)\) is called simple if it is non-trapping, has no conjugate points, and its boundary is strictly convex. Additionally, the authors prove the existence of special solutions to the transport equation called integrating factors. The main novelty in the article is a new construction of a twistor space of \((M,g)\) and a twistor correspondence. The twistor space \(Z\) is a complex surface which is diffeomorphic to the unit disc bundle over \(M\), but the projection to \(M\) is not holomorphic. The complex structure is defined roughly in terms of the vector fields \(X,X_{\perp}\), an extension of \(V\), and the complex coordinate on the disc. In the paper it is observed that \(Z\) satisfies a Grauert-Oka-type principle - for simple \((M,g)\) every holomorphic vector bundle on \(Z\) is trivial. One potential future benefit of this \textit{transport twistor space} is the twistor correspondence established in the paper, which is similar to the known twistor correspondences that appeared after the pioneering work of \textit{R. Penrose} [Rep. Math. Phys., 12(1), 65--76 (1977)]. The correspondence relates transport properties on \(M\) with the complex geometry on \(Z\).