Asymptotically geodesic surfaces

Fernando Al Assal and Ben Lowe · 2025

Read Asymptotically geodesic surfaces by Fernando Al Assal and Ben Lowe (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-17303.txt.

Introduction

Fernando Al Assal and Ben Lowe study surfaces that grow ever flatter in a 2025 math paper. The surfaces sit inside a curved three-dimensional space and close on themselves. A sequence counts when its bending fades evenly toward zero at each step.

When the space has finite volume, such sequences must spread densely through the bundle of flat planes. When volume is endless yet the space stays tame, no such sequence can exist. The split gives partial answers to a wider question on curved spaces that hold flat-like surfaces. Examples in higher dimensions show looser limits there. The work maps where flatness can gather.

Catalog record

CREATORFernando Al Assal and Ben Lowe
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE74 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-17303

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Asymptotically geodesic surfaces Fernando Al Assal and Ben Lowe Abstract . A sequence of distinct closed surfaces in a hyperbolic 3-manifold M M is asymptotically geodesic if their principal curvatures tend uniformly to zero. When M M has finite volume, we show such sequences are always asymptotically dense in the 2-plane Grassmann bundle of M M . When M M has infinite volume and is geometrically finite, we show such sequences do not exist. As an application of the former, we obtain partial answers to the question of whether a negatively curved Riemannian 3-manifold that contains a sequence of asymptotically totally geodesic or totally umbilic surfaces must be hyperbolic. Finally, we give examples to show that if the dimension of M M is greater than 3, the possible limiting behavior of asymptotically geodesic surfaces is less constrained than for totally geodesic surfaces. 1. Introduction A sequence of distinct connected closed surfaces in a hyperbolic 3-manifold M is asymptotically geodesic if their principal curvatures tend to zero in L^{\infty} norm. Kahn-Marković and Kahn-Wright showed that when M has finite volume, such sequences of surfaces exist abundantly [ KM12b ] , [ KW21 ] . The first theorem (Theorem 1.1 ) in this article implies that such sequences (S_{n}) of surfaces are always asymptotically dense in the 2-plane Grassmann bundle \Gr M of M – given an open B\subset\Gr M , there is N so for every n\geq N , S_{n} intersects B . This result is used to obtain partial answers to variations on the question of whether a closed negatively curved Riemannian manifold that contains a sequence of asymptotically geodesic surfaces is hyperbolic (Theorems 1.7 , 1.9 .) We also show (Theorem 1.6 ) that sequences of distinct asymptotically geodesic surfaces do not exist in geometrically finite hyperbolic 3-manifolds M of infinite volume . Precisely, we show there is \epsilon(M)>0 so that a closed surface of principal curvatures smaller than \epsilon in absolute value is homotopic to a totally geodesic surface. There can be real qualitative differences in the behavior of sequences of totally geodesic surfaces and sequences of surfaces that are only asymptotically geodesic. For example, the first author showed that a sequence of distinct asymptotically geodesic surfaces \Sigma_{n} in a hyperbolic 3-manifold can scar along a closed totally geodesic surface \Sigma , in the sense that the corresponding sequence of probability measures \mu_{\Sigma_{n}} converges to the probability measure corresponding to \Sigma [ AA ] . This is in contrast to what happens for sequences of distinct closed totally geodesic \Sigma_{n} in M , which Mozes-Shah showed must become uniformly distributed [ MS95 ] . Most of our theorems are in the opposite direction, and show that asymptotically geodesic surfaces in negative curvature enjoy many of the same rigidity properties as totally geodesic surfaces. In fact, many of our proofs rely on theorems from homogeneous dynamics that establish those rigidity properties for totally geodesic surfaces, such as the fact proven by Ratner and Shah that they are either closed or dense in a hyperbolic 3-manifold of finite volume [ Rat91 ] , [ Sha91 ] . On the other hand, we construct sequences of asymptotically Fuchsian ( K_{n} -quasifuchsian for K_{n}\to 1 ) pleated surfaces and minimal surfaces in certain hyperbolic d -manifolds M with d\geq 4 whose Hausdorff limit is a union of two distinct totally geodesic submanifold of M (Theorem 1.5 .) This contrasts with the fact, that follows from the work of Mozes-Shah, that the Hausdorff limit of a sequence of totally geodesic surfaces in M , if it exists, has to be a single totally geodesic k -submanifold for 3\leq k\leq d (Proposition 5.2 .) 1.1. Density of asymptotically geodesic surfaces in finite volume Let M=\Gamma\backslash\mathbf{H}^{3} be a hyperbolic 3-manifold of finite volume, where \Gamma\leq\PSL_{2}\mathbf{C} is a lattice. All surfaces in this paper will be connected. An essential surface S\subset M is K -quasifuchsian if \pi_{1}(S)\leq\pi_{1}M=\Gamma is a K -quasifuchsian subgroup of \PSL_{2}\mathbf{C} , i.e., there is a K -quasiconformal homeomorphism \phi:\partial_{\infty}\mathbf{H}^{3}\to\partial_{\infty}\mathbf{H}^{3} so \phi\circ\pi_{1}(S)\circ\phi^{-1} is Fuchsian. For K sufficiently close to 1, work by Uhlenbeck and Seppi imply that K -quasifuchsian surfaces S\subset M are homotopic to unique minimal surfaces with principal curvatures going to zero uniformly as K\to 1 [ Uhl83 ] , [ Sep16 ] . All essentials immersed surfaces in this paper are maximal: they are not homotopic to covers of some surface in M of smaller genus. Our first theorem is: Theorem 1.1 . Let M be a finite volume hyperbolic 3-manifold. Let B be an open set. Then, there is \epsilon=\epsilon(B)>0 such that all but finitely many of the (1+\epsilon) -quasifuchsian minimal surfaces of M meet B . Moreover, the ones which do not are totally geodesic. As direct consequences of the previous theorem (and its proof), we obtain the following two corollaries. We say that a surface S is \epsilon -dense in \Gr M if every tangent 2-plane to M is at a distance of at most \epsilon from some tangent plane to S . Corollary 1.2 . For every \epsilon>0 there is some \eta=\eta(M)>0 so that all but finitely many (1+\eta) -quasifuchsian closed minimal surfaces are \epsilon -dense in the \eta -thick part of M . Corollary 1.3 . Let (S_{n}) be a sequence of distinct closed asymptotically Fuchsian minimal surfaces in M . Then, (S_{n}) is asymptotically dense in \Gr M : for every open U\subset\Gr M , there is N so that for every n\geq N , S_{n} intersects U . These theorems imply that, in the topology induced by the Hausdorff distance between closed sets in \Gr M , sequences of asymptotically geodesic surfaces S_{n} always limit to all of \Gr M . This is in contrast with the fact that sequences of asymptotically geodesic surfaces S_{n} may not equidistribute in \Gr M – in fact, work by the first author showed that the probability measures \nu_{n} induced by them on \Gr M may limit in the weak-* topology to a measure supported on a single totally geodesic surface in M [ AA ] . A crucial ingredient in these proofs is the Ratner-Shah theorem stating that totally geodesic surfaces are either closed or dense in M . When M has infinitely many distinct totally geodesic surfaces, the fact due to Mozes-Shah that they are equidistributing also plays an important role [ MS95 ] . Remark 1.4 . Results similar to the previous theorem and corollaries were recently and independently obtained by Xiaolong Hans Han [ Han25 ] . 1.2. Hausdorff limits in higher dimensions We say a sequence of surfaces (S_{n}) in a hyperbolic d -manifold M is asymptotically Fuchsian if the S_{n} are K_{n} -quasifuchsian for K_{n}\to 1 as n\to\infty . (Analogously to the 3-dimensional case, we say an essential surface S\subset M is K -quasifuchsian if \pi_{1}(S)\leq\SO(d,1) is a K -quasifuchsian subgroup, meaning there is a K -quasiconformal homeomorphism of \partial_{\infty}\mathbf{H}^{d} conjugating \pi_{1}(S) into \SO(2,1) .) Asymptotically Fuchsian surfaces in M are always homotopic to asymptotically geodesic ones [ Jia21 ] . Conversely, asymptotically geodesic surfaces are asymptotically Fuchsian [ Eps86 ] . We show Theorem 1.5 . Suppose M is a closed hyperbolic d -manifold containing two closed totally geodesic 3-dimensional

Read the complete text