A Remark on the reach and upper bounds on some extrinsic geometry invariants of submanifolds
R. Mirzaie · 2025
Read A Remark on the reach and upper bounds on some extrinsic geometry invariants of submanifolds by R. Mirzaie (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-2503-07054.txt.
Introduction
Reza Mirzaie measures how far a surface sits from trouble inside a curved space. His 2025 paper studies the reach of a compact submanifold. Reach is roughly the largest tube radius around it before the tube folds into itself. It caps how curved and how close to itself the surface can be.
Working in a general curved space rather than a flat one, he uses the second variation formula to tie reach to the surface's outer geometry. That formula tests whether paths are truly shortest. The results relate reach bounds to extrinsic invariants of the embedding. Earlier theorems of this kind were proved only for Euclidean surroundings; Mirzaie's versions carry them into curved ambient spaces, widening where the estimates can be used.
Catalog record
| CREATOR | R. Mirzaie |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-07054
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
A Remark on the reach and upper bounds on some extrinsic geometry invariants of submanifolds Thanks: R. Mirzaie, Department of Pure Mathematics, Faculty of Science, Imam Khomeini International University, Qazvin, Iran [email protected] R. Mirzaie Abstract . We consider a compact submanifold M M of a Riemannian manifold N N and we use the second variation formula as a tool to drive some geometric results on r e a c h ( M , N ) reach(M,N) the reach of M M in N N , including some useful relations between the extrinsic geometry of M M in N N and r e a c h ( M , N ) reach(M,N) . Our results generalize some theorems previously proved for the special case where N N is Euclidean space. Key words: Riemannian manifold, Reach, Submanifold. MSC: : 53C70, 53C22, 53B25, 62C20, 68U05. 1. Introduction The reach of a subset M of Euclidean space was introduced by Federer [9] as the largest number \tau such that any point at distance less than \tau from M has a unique nearest point on M ( foot point on M ). The reach appears under various names in the literature. It was called the condition number in topological data analysis. In computational geometry it is used under the name of local feature size. It is an important geometric invariant of subsets of Euclidean space which plays effective roles in the field of manifold learning and topological data analysis (see [4, 8, 13]) and has been widely used by the researches in geometric measure theory, integral geometry ([10, 15]), image processing arguments and biomathematics. If M is a submanifold of Euclidean space, it is proved that the reach quantifies the maximum ratio between the geodesic distance on M and Euclidean distance for pairs of points in M , and it is a measure of a manifolds departure from convexity. In fact, the reach is a global invariant incorporating informations about the second fundamental form of the embedding and the location of the first critical point of the distance from the submanifold. Thus, it incorporates the local and global topology and geometry of M . Positivity of the reach for subsets of Euclidean space is the minimal regularity assumption on sets. If M is a smooth compact submanifold of Euclidean space, it has a positive reach and if the reach is bounded, then strong topological and geometric restrictions on M will appear. For instance, it is proved that for compact submanifolds of R^{n} with the bounded reach, the angle between tangent spaces at different points are bounded. There are many papers about the relations between the reach of M in N and the geometry and topology of M when N is a Euclidean space. We refer to [7] and [2] for some results about the relations between the reach and the geometry of M when N is Euclidean space or it is a sphere. Authors of [11] and [5] extended the reach from subsets of Euclidean space to the reach of subsets of Riemannian manifolds. Although some results about the reach obtained for the submanifolds of Euclidean space can directly be extended to the general case where the ambient space is a Riemannian manifold, but in many cases we need more calculations and differential geometric tools. In the present article, we consider a compact differentiable submanifold M of a Riemannian manifold N . In Theorem 3.2, we find an upper bound for the exterior part of the acceleration vector of the unit speed geodesics of M , which is a calculation of the impact of the reach on the exterior geometry of M . Then, we drive some more results about the geometry of M and as application we use our theorem in the special case where N has constant curvature. 2. Preliminaries The main tool in our proofs is the second variation formula which measures the change in the arch length under the small displacements. We refer to [14] for definitions and further results. In what follows, N is a smooth and complete Riemannian manifold and M is a connected compact submanifold of N . All maps are considered to be smooth. Remark 2.1 (Second variation formula ). Let \sigma:[0,1]\to N be a geodesic segment and \chi:[0,1]\times(-\delta,\delta)\to N be a smooth map such that \chi(t,0)=\sigma(t) . Then, \chi is called a variation of \sigma . Denote by L(s) the length of the curve \sigma_{s}:[0,1]\to N , t\to\chi(t,s) ( \sigma_{0}=\sigma ) and let V(t)=\frac{d}{ds}\chi(t,s)_{{}_{|s=0}}=\chi_{{}_{s}}(t,0) , A(t)=\chi_{{}_{ss}}(t,0) . By the first and second variation formulas we have: L^{\prime}(0)=\frac{1}{L(0)}\int_{0}^{1}<\sigma^{\prime}(t),V^{\prime}(t)>dt and L^{\prime\prime}(0)=\frac{1}{L(0)}\int_{0}^{1}\{<V^{\prime\perp},V^{\prime\perp}>-<R_{V\sigma^{\prime}}V,\sigma^{\prime}>\}dt +\frac{1}{L(0)}(<\sigma^{\prime}(1),A(1)>-<\sigma^{\prime}(0),A(0))>). Where, (V^{\prime})^{\perp} is the component of V^{\prime} normal to \sigma^{\prime} . Similarly, if for all s\in(-\delta,\delta) , the curve \sigma_{s} is a geodesic in N , then L^{\prime}(s)=\frac{1}{L(s)}\int_{0}^{1}<\sigma_{s}^{\prime}(t),V_{s}^{\prime}(t)>dt L^{\prime\prime}(s)=\frac{1}{L(s)}\int_{0}^{1}\{<V_{s}^{\prime\perp},V_{s}^{\prime\perp}>-<R_{V_{s}\sigma_{s}^{\prime}}V_{s},\sigma_{s}^{\prime}>\}dt +\frac{1}{L(s)}(<\sigma_{s}^{\prime}(1),A_{s}(1))>-<\sigma_{s}^{\prime}(0),A_{s}(0)>). where, V_{s}(t)=\chi_{s}(t,s) , A_{s}(t)=\chi_{ss}(t,s) . Definition 2.2 (medial axes and reach). Let M be a compact submanifold of a Riemannian manifold N . 1. The medial axes of M in N which we denote by med(M,N) is the closure of the following subset of N . A=\{x\in N:\exists p\neq p^{\prime}\in M\ such\ that\ d(x,M)=d(x,p)=d(x,p^{\prime})\} 2. The reach of M in N which we denote by \tau , is defined as follows: \tau=inf\{d(x,M):x\in med(M,N)\}. 3. If q\in N , then a point p\in M , with d(q,M)=d(q,p) is called a foot point of q . 4. A point q\in med(M,N) is called a reach assigning point of Med(M,N) if d(q,M)=\tau . A foot point of a reach assigning point of Med(M,N) is called a reach assigning point on M . Remark 2.3. If N is compact then there is a reach assigning point on med(M,N) . In fact, by definition of the reach, for each positive integer n , there is a point q_{n}\in Med(M,N) such that \tau\leq d(q_{n},M)<\tau+\frac{1}{n} . By the compactness of N , \{q_{n}\} has a convergent subsequence which we denote it again by \{q_{n}\} . If q=lim_{n\to\infty}q_{n} , then d(q,M)=\tau . If p\in M , denote by (T_{p}M)^{\perp} the subspace of all vectors normal to M at p . Consider the usual exponential map exp_{p}:T_{p}N\to N , and let B_{r}(p)=exp_{p}(\{v\in(T_{p}M)^{\perp}:|v|\leq r\}) . Remark 2.4. If reach(M,N)=\tau , then for all positive numbers r<\tau , the r -tube around M , defined by \{x\in N:d(x,M)<r\} is an embedded submanifold of N ( a consequence of the tubular neighborhood theorem). Consequently, the reach of M in N can also be defined as: sup\{\eta>0:B_{\eta}(p)\cap B_{\eta}(p^{\prime})=\emptyset,for\ all\ p,p^{\prime}\in M\ with\ p\neq p^{\prime}\}. 3. Some inequalities describing the impact of the reach on the extrinsic geometry of submanifolds It is proved in [7] that the reach gives an upper bound for the length of the acceleration of the arc length parametrized geodesics in the compact submanifolds of the Euclidean space. Also we refer to [1] for more similar calculations relating the reach and the extrinsic geometry of the submanifolds of Euclidean space. The theorems of the present section can be used as a tool to obtain some similar results for the compact submanifolds of a Riemannian manifold. In what follows, we will denote by \bar{\nabla} the Levi Civita connection of N . \nabla will denote to the induced