Multidimensional Scaling of Varietal Data in Sedimentary Provenance Analysis

Abstract

is a multivariate ordination technique that has gained considerable popularity in recent years as a method to interpret large data sets in sedimentary provenance analysis (Vermeesch, 2013). Given a table of pairwise "dissimilarities" between samples, MDS produces a lower (typically two-) dimensional "map" in which similar samples plot close together and dissimilar samples plot far apart. MDS can be applied to a wide variety of different provenance proxies by choosing an appropriate dissimilarity measure. Vermeesch (2019) distinguishes between three different types of provenance data, each of which is associated with a specific database format (Figure 1):1. Distributional data such as detrital zircon U-Pb ages can be stored in lists of decimal numbers, where each list represents a sample and typically contains a different number of values (i.e., single grain ages). Two samples can be compared using the Kolmogorov-Smirnov distance or related nonparametric statistics (Vermeesch, 2018a). The resulting dissimilarity matrix fulfills the metric requirements and is therefore suitable for both classical and nonmetric MDS (Vermeesch, 2013).