Quantifying local anisotropy in point pattern analysis
Conference
65th ISI World Statistics Congress
Format: IPS Abstract - WSC 2025
Keywords: spatial statistics
Session: IPS 933 - Spatial Linear Networks and Directional Analysis
Wednesday 8 October 10:50 a.m. - 12:30 p.m. (Europe/Amsterdam)
Abstract
A spatial point pattern is termed anisotropic if its statistical properties exhibit directional variation. Global anisotropy, which occurs when a spatial point pattern exhibits uniform directional characteristics across the study area, has been extensively studied. Local anisotropy analysis for point patterns is less commonly addressed in the literature and often developed for particular applications. In many real-world applications, anisotropy may vary spatially due to environmental, structural, or anthropogenic factors. Thus, accurately characterising point patterns with heterogeneous directional behaviour requires understanding and quantifying local anisotropy. Herein, we develop a methodology to quantify local anisotropy, providing spatially resolved measures of direction and strength across the point pattern domain.