Interpretation of the Richardson plot in time series representation

摘要

In fractal analysis of a time series, the relationship between series length and ruler length may be represented graphically as a Richardson Plot. Fractal dimension measures can be estimated for particular ranges of ruler length, supported by the graphical representation. This paper discusses Richardson Plots which have been obtained for several types of time series. From these, patterns have been identified with explanations. There is particular focus on local maxima and minima. Significant influences found present are described as gradient and vertex effects. The task - and implications - of partitioning the range of ruler lengths in determining fractal dimension measures is briefly addressed.