The use of kernel density estimates on chemical and isotopic data in archaeology

Abstract

• Kernel Density Estimates (KDEs) have found wide application in archaeological spatial analysis, where they have been integrated into geographical information systems with various user-friendly packages to explore the spatial distribution of archaeological finds. The creation of a KDE provides a non-parametric way of converting continuous data such as histograms or scatter plots into a smoothed probability density function, which allows inferences to be made about the population distribution based on a given sample of data. The open-access software, developed by Christopher Bronk Ramsey for producing and comparing KDE distributions of glass chemistry in volcanic tephra, can be used to generate KDEs and test for overlaps. The earliest application of KDE in archaeological chemistry was in relation to the interpretation of lead isotope data. Trace element analysis can be used to infer the source of the raw materials used to produce glass.