Face volume densities of positive-intensity and ideal Poisson–Voronoi tessellations in hyperbolic spaces

Matteo D’Achille & Christoph Thäle · 2026

Read Face volume densities of positive-intensity and ideal Poisson–Voronoi tessellations in hyperbolic spaces by Matteo D’Achille & Christoph Thäle (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-2606-26049.txt.

Introduction

Matteo D'Achille and Christoph Thäle compute face statistics for Poisson-Voronoi tessellations in hyperbolic space. A Voronoi tessellation carves space into cells around random seed points. Each cell holds the spots closest to its seed. Hyperbolic space curves negatively, which changes the geometry throughout. Earlier exact results covered only dimensions two and three, in work by Isokawa.

Formulas now cover every dimension and every face size. The authors determine the volume densities of faces of each dimension for any positive seed intensity. A new formula of Blaschke-Petkantschin type, in hyperbolic space, does the heavy lifting. The ideal tessellation, reached as intensity shrinks toward zero, gets closed forms too. All face volume densities and typical face volumes come out explicitly there. The results extend a classical strand of stochastic geometry into curved space.

Catalog record

CREATORMatteo D’Achille & Christoph Thäle
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE46 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2606.26049

Browse the full library · About the Tower of Babel

Full text

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

Face volume densities of positive-intensity and ideal Poisson–Voronoi tessellations in hyperbolic spaces Matteo D’Achille & Christoph Thäle Abstract We determine analytically for all k ∈ { 0 , 1 , … , d − 1 } k\in\{0,1,\ldots,d-1\} the k k -volume densities of a Poisson–Voronoi tessellation of intensity λ > 0 \lambda>0 in the d d -dimensional hyperbolic space of constant curvature − 1 -1 . This largely extends previous results of Isokawa in dimensions two and three. As applications, we provide closed form expressions for all face volume densities and all typical face volumes of the ideal Poisson–Voronoi tessellation (IPVT), which is the low-intensity limit as λ ↓ 0 \lambda\downarrow 0 of the hyperbolic Poisson–Voronoi tessellation. As a main tool we develop a new Blaschke–Petkantschin–type formula in hyperbolic space. Keywords: Blaschke–Petkantschin formula; hyperbolic geometry; ideal Poisson–Voronoi tessellation (IPVT); k k -faces; Poisson–Voronoi tessellation MSC (2020): 52A55, 60D05, 60G55 1 Introduction and main result Poisson–Voronoi tessellations and Blaschke–Petkantschin formula in Euclidean spaces. Poisson–Voronoi tessellations are among the most classical models of random spatial subdivision. Starting from a stationary Poisson point process in Euclidean space, one assigns to each point the region consisting of all locations for which this point is the nearest point of the process. This construction produces a random tessellation whose cells describe the domains of influence of the points of the process. Due to its simple definition and rich geometric structure, the Poisson–Voronoi tessellation has become a standard model and has found numerous applications, for instance in materials science, telecommunications, geographical modelling and the analysis of spatial data. Several fundamental mean values of the Euclidean Poisson–Voronoi tessellation in \mathbb{R}^{d} are known analytically. These include, in particular, the counting densities of k -faces and the corresponding k -volume densities for all k\in\{0,1,\ldots,d-1\} , see [ 26 , 27 , 30 ] and the references mentioned therein. Such formulas are basic quantitative descriptors of the tessellation. They determine how many faces of a given dimension occur per unit volume and how much k -dimensional content these faces carry on average. They therefore provide a benchmark for simulations, a point of comparison between different random tessellation models, and a starting point for the study of more refined characteristics. A central role in the derivation of such formulas is played by Blaschke–Petkantschin–type transformations involving spheres. These integral-geometric change-of-variables formulas allow one to separate location, radius and directional information in configuration integrals, and thereby turn geometric characteristics of the tessellation into tractable lower-dimensional integrals. Beyond Poisson–Voronoi tessellations, Blaschke–Petkantschin formulas of this type have become versatile tools in stochastic and computational geometry. We refer to the survey [ 28 ] , and mention selected applications in [ 6 , 10 , 11 , 12 , 29 ] . Poisson–Voronoi tessellations and Blaschke–Petkantschin formula in hyperbolic spaces. It is natural to ask to what extent the analytic theory for Poisson–Voronoi tessellations survives in non-Euclidean geometries. In the present paper we focus on hyperbolic spaces of constant negative curvature -1 . Hyperbolic Poisson–Voronoi tessellations retain the basic nearest-neighbour interpretation of their Euclidean counterparts, but the underlying geometry becomes substantially different. In particular, the exponential growth of volume and the absence of a scaling as in Euclidean spaces significantly influence the topology and geometry of the cells, and make explicit computations far more delicate. Compared with the Euclidean case, considerably fewer characteristics of hyperbolic Poisson–Voronoi tessellations are known in closed form. Detailed studies in dimensions two and three were carried out in [ 20 , 19 ] , where mean values and structural properties of the corresponding tessellations were investigated based on hyperbolic trigonometry, which is possible only in these low-dimensional cases. In general dimension, first explicit progress was made in [ 15 ] . The approach there is based on the relation of hyperbolic Poisson–Voronoi tessellations to so-called beta-star polytopes and yields for any dimension d\geq 2 and any k\in\{0,1,\ldots,d-1\} an explicit description of the expected number of k -faces of the typical cell. These numbers determine in turn the k -face counting densities of the hyperbolic Poisson–Voronoi tessellation. However, these results only concern the combinatorial structure of the tessellation. They do not determine the metric size of its lower-dimensional skeletons. In particular, no formula for the k -volume density of the hyperbolic Poisson–Voronoi tessellation in arbitrary dimension seems to have been available so far. The present paper studies the complementary metric quantity, namely the k -volume density of the hyperbolic Poisson–Voronoi tessellation, in arbitrary dimension. Thus, instead of counting k -faces, we measure their total k -dimensional volume per unit hyperbolic volume. This provides the metric counterpart to the face-counting formulas obtained in [ 15 ] . It also substantially extends the explicit computations in low dimensions from [ 20 , 19 ] and complements the asymptotic local results of [ 4 ] for the face-counting densities of Poisson–Voronoi tessellations on general Riemannian manifolds. The relevance of such metric characteristics is already visible in dimension two: the low-intensity limit of the length density of the one-dimensional skeleton, equivalently the k -volume density for k=1 , plays a central role in the upper bound for the Cheeger constant of high genus hyperbolic surfaces obtained in [ 2 ] . This connection will be discussed further below in relation to the ideal Poisson–Voronoi tessellation. A key ingredient in our approach is a new Blaschke–Petkantschin–type formula in hyperbolic space. It decomposes configuration integrals according to the geodesic subspace, the centre and the radius of the circumsphere determined by a collection of points. A boundary case of our formula, corresponding to d+1 points with a unique hyperbolic circumcentre, was previously obtained in [ 5 , Proposition 2.1.1] . The form needed here extends this decomposition to lower-dimensional circumspheres and is therefore suited to the analysis of k -faces for every k\in\{0,1,\ldots,d-1\} . Since Euclidean Blaschke–Petkantschin formulas of this type have proved to be a versatile tool in the applications discussed above, it is natural to expect that a hyperbolic counterpart will be useful beyond the specific computation carried out here. We therefore regard this formula as an object of independent interest. We moreover mention that the ratio of face-counting densities and face-volume densities available from [ 20 ] has been used in [ 17 ] to compute the right derivative at \lambda=0 of the critical probability p_{c}(\lambda) for Poisson–Voronoi percolation on the hyperbolic plane, where the percolation model is defined by colouring independently the cells of a Poisson–Voronoi tessellation with intensity \lambda . We leave the investigation of the implications of the present work for hyperbolic Poisson–Voronoi percolation in dimensions d\geq 3 to future work. Figure 1.1 : Left: sample of the Poisson–Voronoi tessellation of \mathbb{H}^{2} in the conformal disk model at positive intensity (nuclei in red). Right: sample of {\rm IPVT}(\mathbb{H}^{2}) coupled via dilation from \mathbf{o} . In both figures, only the first 10000 nuclei have been used (and displayed in the left one). Ideal Poisson–Voronoi tessellations. The explicit formula for the k -volume density of a hyperbolic Poisson–Voronoi tessellation we prove makes it possible to study the low-intensity regime. This connects the present work to the ideal Poisson–Voronoi tessellation (IPVT), a recently introduced limiting object obtained from hyperbolic Poisson–Voronoi tessellations when the number of nuclei per

Read the complete text