The Linear Spatial-Temporal Interaction Process and its Relation to 1/ω-Noise
Journal of the Royal Statistical Society: Series B (Statistical Methodology) · 1994 · Volume 56 · Issue 1
DOI: 10.1111/j.2517-6161.1994.tb01961.x
Abstract
If we are to gain an understanding of the evolutionary processes which generate pattern then time cannot be ignored, since a purely spatial viewpoint disregards the fundamental generating mechanism of the underlying process. We therefore develop a spectral approach based on a general growth–interaction model; individuals grow at intrinsic rateλ, while two individuals a distancerapart interact at rateartimes their size difference. Special cases considered involve nearest neighbour, geometric, logarithmic and Cauchy-type dependence. A separate study is made of ‘long-term persistence‘; an autoregressive binomial weight structure is used to generate processes with power law spectraωd.We examine the effect of arranging the general process on a ring ofNsites, and then use this structure to simulate nearest neighbour and binomial interaction.
Keywords
competition, fourier transform, fractional brownian motion, growth, interaction, long range dependence, nearest neighbour interaction, ω-stationarity, pattern, power law, ring process, simulation, spatial–temporal process, spectrum
Enable JavaScript for citation tools and additional paper interactions.