Diffusion Smoothing for Spatial Point Patterns

Abstract

Traditional kernel methods for estimating the spatially-varying density of points in a spatial point pattern may exhibit unrealistic artefacts,in addition to the familiar problems of bias and over or under-smoothing.Performance can be improved by using diffusion smoothing, in which thesmoothing kernel is the heat kernel on the spatial domain. This paper developsdiffusion smoothing into a practical statistical methodology for twodimensionalspatial point pattern data. We clarify the advantages and disadvantagesof diffusion smoothing over Gaussian kernel smoothing. Adaptivesmoothing, where the smoothing bandwidth is spatially-varying, can beperformed by adopting a spatially-varying diffusion rate: this avoids technicalproblems with adaptive Gaussian smoothing and has substantially betterperformance. We introduce a new form of adaptive smoothing using laggedarrival times, which has good performance and improved robustness. Applicationsin archaeology and epidemiology are demonstrated. The methods areimplemented in open-source R code