Planar geodesic immersions in pseudo-Euclidean space (Q1062297): Difference between revisions

An isometric immersion \(f: M^ n_ r\to {\mathbb{R}}^ N_ s\) is said to be planar geodesic if the image of each geodesic of M lies in a 2-plane of \({\mathbb{R}}^ N_ s\). Planar geodesic submanifolds of Euclidean space were studied and classified by Hong, Little, Sakamoto, and Ferus; they are spheres and certain Veronese imbeddings of projective spaces. In the present work, this problem is considered for indefinite metrics. As in the positive-definite case, the length of the second fundamental form is constant, but here the immersion is not necessarily parallel. Among other results, the following list is given of all parallel, planar geodesic, non-totally geodesic, isometric immersions on surfaces in \({\mathbb{R}}^ N_ s:\) the standard imbeddings of pseudo-Riemannian spheres \(S^ 2_ r\) and \(H^ 2_ r\) in \({\mathbb{R}}^ 3_ s\), the Veronese immersions of \(S^ 2_ r\) and \(H^ 2_ r\) in \({\mathbb{R}}^ 5_ s\), and flat surfaces of the form \[ (x,y)\in {\mathbb{R}}^ 2_ r \to (g_ 1(x,y),...,g_ k(x,y),\quad x, y, g_ 1(x,y),...,g_ k(x,y))\in {\mathbb{R}}_{r+k}^{2+2k}, \] with \(k=1,2\), or 3, where the metric on \({\mathbb{R}}_{r+k}^{2+2k}\) has signature (-,..,-,\(+,..,+)\) and \(g_ i\), \(i=1,...,k\), are quadratic polynomials.