Application of multivariate statistical techniques in alongshore differentiation of coastal barriers

This paper investigates the existence of statistical correlations and relationships between geological and oceanographic parameters along a coastal barrier system in southern Brazil. The existence of few statistical studies on coastal barriers motivated the use of multivariate statistical techniques that take environmental parameters such as coastal morphological-sedimentary features and their morphometric measurements into account. The Hierarchical Cluster Analysis (HCA) and Principal Component Analysis (PCA) were used to assess the spatial variations of fourteen parameters driving the current and past behavior of this coastal barrier system at fourteen locations. Through the application of HCA, these locations were divided into four groups (Groups 1, 2, 3, and 4). The HCA results were analyzed using PCA. The first (PC1) and the second (PC2) principal components together explained at least 76.4% of the variability of all four groups. The most important geological and oceanographic parameters that were correlated to PC1 and PC2 were different in each group. In general, the shoreface slope and shoreface dimensions were common to all groups as PC1-correlated parameters, being the most important ones regarding data variability. In addition, for the progradational (Groups 1 and 4), retrogradational (Group 2) and aggradational barriers (Group 3) PCA indicated that shoreline orientation (wave exposure) and beach type are secondary, but relevant factors controlling the past and current behavior of coastal barriers in southern Brazil.