High-intensity convergence to the parabolic hull tessellation

Determine whether, under the high-intensity scaling \(\gamma=\lambda^{d+1}\) as \(\lambda\to\infty\), the transformed downward-paraboloid process associated with the hyperbolic horoball model converges to the tessellation induced by the parabolic hull process.

Background

The paper establishes the low-intensity scaling limit γ=λd1\gamma=\lambda^{d-1} and proves convergence of the associated projected Poisson–Laguerre tessellations to a classical Poisson–Delaunay tessellation. In contrast, the authors identify a different high-intensity regime, γ=λd+1\gamma=\lambda^{d+1}, which is expected to recover the Euclidean parabolic-hull geometry associated with random polytopes near a smooth boundary.

The authors provide limiting cell-intensity constants in this high-intensity regime, including c2=(2/3)1/3Γ(5/3)c_2=(2/3)^{1/3}\Gamma(5/3) and c3=35/24c_3=35/24, but do not establish convergence of the underlying transformed process or tessellation. Thus, proving the stated process-level convergence remains unresolved in the paper.

References

More precisely, if one sets $\gamma=\lambda{d+1}$, then, as $\lambda\to\infty$, the transformed downward-paraboloid process is expected to converge to the tessellation induced by the parabolic hull process. We do not pursue this limit here.

Random hyperbolic polyhedra in horoballs  (2609.10007 - Besau et al., 9 Sep 2026) in Section “The cell intensity,” remark immediately following Theorem \(\ref{thm:AsymptoticsGeneralD}\)