Overview of Numerical Methods in Palaeolimnology

Abstract

An overview of the major numerical methods for the analysis of modern-surface-sample and stratigraphical data is given in this chapter. The methods are discussed in relation to data collection and assessment, data summarisation, data analysis, and data interpretation. The main methods and research approaches outlined include data storage and data-bases, exploratory data analysis, clustering and partitioning, classical and canonical ordination, partial ordinations, principal curves, classification and regression trees and related decision-tree methods, artificial neural networks and self-organising maps, estimating gradient lengths, estimating taxonomic richness, estimating species optima and tolerances, Procrustes rotation for comparing ordinations, and variation partitioning. Many of these techniques are also discussed in detail in some of the chapters in Parts III and IV.