Urysohn width and macroscopic scalar curvature

Aditya Kumar, Balarka Sen · 2026

Read Urysohn width and macroscopic scalar curvature by Aditya Kumar, Balarka Sen (2026) 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-2601-14669.txt.

Introduction

Aditya Kumar and Balarka Sen disprove a conjecture in geometry in this 2026 paper. The topic is Urysohn width, a measure of how well a space can be squeezed onto a lower-dimensional skeleton. Gromov proposed a macroscopic version of his Urysohn width conjecture for scalar curvature. The authors show that version is false in dimensions four and above.

Their proof rests on two parts. One is a novel estimate on the codimension two Urysohn width of circle bundles over manifolds with large hypersphericity radius. The other is a notion of ruling for Riemannian manifolds. The ruling yields circle bundles whose total spaces admit metrics of positive macroscopic scalar curvature. Readers get a clean counterexample in a deep area.

Catalog record

CREATORAditya Kumar, Balarka Sen
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE75 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-14669

Browse the full library · About the Tower of Babel

Full text

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

Urysohn width and macroscopic scalar curvature Aditya Kumar, Balarka Sen Abstract . We show that the macroscopic version of Gromov’s Urysohn width conjecture for scalar curvature is false in dimensions four and above. This is based on (1) a novel estimate on the codimension two Urysohn width of circle bundles over manifolds with large hypersphericity radius, and (2) a notion of ruling for Riemannian manifolds that yields circle bundles with total spaces admitting metrics of positive macroscopic scalar curvature. Along the way, we also show that Urysohn width is not continuous under Cheeger-Gromov collapsing limits. This article is a continuation of our study of metric invariants and scalar curvature for circle bundles over large Riemannian manifolds initiated in [ 28 ] . 1. Introduction 1.1. Background and motivation Definition 1.1 (Urysohn width) . Let X be a metric space. The Urysohn k -width of X , denoted by \mathrm{UW}_{k}(X) , is the infimum over all t>0 such that there is a continuous map f:X\to K to a simplicial complex K of dimension at most k satisfying the property that the diameter of each fiber of f is at most t , i.e., for all p\in K , \mathrm{diam}(f^{-1}(p))\leq t . The notion of Urysohn width originated in the works of Pavel Urysohn on topological dimension theory [ 37 , 1 ] . Heuristically, \mathrm{UW}_{k}(X)\leq\varepsilon indicates that the metric space X looks like a k -dimensional simplicial complex upon ignoring features at scale \varepsilon or less. In Riemannian geometry, the notion of Urysohn width was popularized by several works of Gromov [ 19 , 18 ] . It arises naturally in the study of non-collapsing and thick-thin decompositions and has also played an important role in systolic geometry. Some of the recent works on these topics have been [ 25 , 33 , 34 , 35 , 6 , 2 ] . For a comprehensive treatment of Urysohn width, we refer the reader to the thesis of Balitskiy [ 7 ] . Definition 1.2 (Scalar curvature) . Let (M^{d},g) be a Riemannian manifold. The scalar curvature of M at a point x , denoted by \mathrm{Sc}_{M}(x) , is the coefficient appearing in the (d+2) th order Taylor expansion of the volume of an infinitesimal ball centered at x : \mathrm{vol}_{M}(B_{r}(x))=\omega_{d}r^{d}\left(1-\frac{\mathrm{Sc}_{M}(x)}{6(d+2)}r^{2}+O(r^{3})\right), where \omega_{d} denotes the volume of a unit ball in the Euclidean d -space. The scalar curvature can be computed infinitesimally as the trace of the Ricci curvature tensor \mathrm{Sc}(x)=\mathrm{tr}~\mathrm{Ric}_{x} . The study of the geometry and topology of manifolds possessing positive scalar curvature metrics has seen an explosion of activity in recent years. The primary motivation for this article is the following outstanding conjecture of Gromov on the Urysohn width of Riemannian manifolds with positive scalar curvature (see [ 18 , 2.A] , [ 19 , G4] , [ 16 , 2½] ). Conjecture 1 (Gromov) . Let M^{d} be a complete Riemannian manifold of dimension d . If M admits a metric with scalar curvature \mathrm{Sc}\geq\sigma^{2}>0 , then there exists a dimensional constant C_{d} such that \mathrm{UW}_{d-2}(M)\leq C_{d}/\sigma . The primary evidence in this direction is the following simple observation: Example 1.1 . Given any closed Riemannian manifold M^{d-2} , consider its product with a round 2 -sphere of radius \varepsilon , i.e., the manifold X^{d}=M^{d-2}\times S^{2}({\varepsilon}) equipped with the product metric. Then, for \varepsilon>0 sufficiently small, the scalar curvature of X satisfies: \mathrm{Sc}_{X}=\mathrm{Sc}_{M}+2\varepsilon^{-2}\geq\varepsilon^{-2}. Fibers of the projection X\to M have diameter \mathrm{diam}(S^{2}(\varepsilon))=\pi\varepsilon . Thus, \mathrm{UW}_{d-2}(X)\lesssim\varepsilon. Conjecture 1 asserts that every complete manifold with a uniform positive lower bound on the scalar curvature exhibits the same behavior as Example 1.1 . One could view Conjecture 1 as the assertion that if the volume of infinitesimal balls in (M,g) is less than that of Euclidean balls of the same radius, then the codimension 2 Urysohn width of M is uniformly bounded. So far, Conjecture 1 is only known to hold in dimension 3 [ 31 , 11 ] . In fact, even the codimension 1 width case involving a uniform upper bound on \mathrm{UW}_{d-1}(M^{d}) remains unknown. If instead of an upper bound on the volume of balls with infinitesimal radius , one has an upper bound on the volume of balls of unit radius , then this provides a large scale or macroscopic notion of positivity of scalar curvature. The following remarkable result due to Guth [ 24 , 25 ] (see also, [ 34 , 33 ] ) deduces a uniform upper bound on \mathrm{UW}_{d-1}(M) if M has sufficiently positive macroscopic scalar curvature, i.e., the volume of every unit ball in M is bounded above by a sufficiently small dimensional constant. Theorem 1.1 (Guth) . There exists a constant \delta_{d}>0 such that for any closed Riemannian manifold (M^{d},g) , if every ball of radius 1 in M has volume at most \delta_{d} , then \mathrm{UW}_{d-1}(M)\leq 1 . The notion of macroscopic scalar curvature and its analogy with scalar curvature was discussed at length by Guth in his ICM address [ 22 , Section 7] . In particular, he states the metaphor that “the macroscopic scalar curvature is like the scalar curvature.” [ 22 , Metaphor 3, pg. 11] . In consonance with this metaphor, in the past few years, there have been several works resolving macroscopic analogs of known as well as conjectural results for scalar curvature [ 23 , 3 , 5 , 10 , 4 , 12 ] . Nonetheless, as observed in [ 2 ] , a straightforward generalization of Theorem 1.1 to codimension 2 Urysohn width does not hold. We give two examples to illustrate this. Example 1.2 . Let S^{d}(r) denote a d -sphere with the round metric of radius r . Let \varepsilon>0 be very small and R\gg 1 be very large. (1) Consider the manifold S^{1}(\varepsilon)\times S^{2}(R) , with the product metric. Unit balls of this manifold are roughly isometric to S^{1}(\varepsilon)\times D^{2}(1) . Hence, they have volume \sim O(\varepsilon) . However, by the fiber contraction lemma (Lemma 2.1 ), \mathrm{UW}_{1}(S^{1}(\varepsilon)\times S^{2}(R))\sim R . (2) Consider S^{3} as the total space of the Hopf fibration S^{1}({\varepsilon})\to S^{3}\xrightarrow{\pi}S^{2}(R) equipped with the Berger metric g_{\varepsilon,R}=\pi^{*}g_{S^{2}(R)}+\varepsilon^{2}\theta\otimes\theta . Once again, unit balls of (S^{3},g_{\varepsilon,R}) are roughly isometric to S^{1}(\varepsilon)\times D^{2}(1) . Therefore, they have volume \sim O(\varepsilon) . Since \pi is not nullhomotopic, the fiber contraction lemma implies \mathrm{UW}_{1}(S^{3},g_{\varepsilon,R})\sim R . Note that as scalar curvature is an infinitesimal quantity, a bound of the form \mathrm{Sc}\geq\sigma^{2} on (M,g) continues to hold upon passage to the universal cover (\widetilde{M},\widetilde{g}) . The first example is eliminated by imposing a control on the volumes of balls in the universal cover. Indeed, the universal cover of S^{1}(\varepsilon)\times S^{2}(R) is \mathbf{R}\times S^{2}(R) , and unit balls here have substantial volume. However, it does not eliminate the second example, as S^{3} is simply connected. To remedy this, Alpert-Balitskiy-Guth [ 2 ] propose that one should impose a control on the volume of unit balls in the universal cover of a slightly larger concentric ball. We reproduce their proposal as a definition below. Note that the following definition slightly differs from the notion of macroscopic scalar curvature defined in [ 22 , Section 7] . Definition 1.3 (Macroscopic scalar curvature) . Let (M^{d},g)

Read the complete text