The present work attempts to describe the spatial distribution of heavy met
als in the surficial sediments of the Kattegat and Skagerrak. It is based o
n multivariate statistical analyses. Principal Component Analysis is used t
o gain information on the chemical characteristics of the metals as a group
. Cluster analysis is used to classify the sampling stations into groups re
garding the principal components. In addition, cluster analysis is adopted
directly on the metals. Together these enable an interpretation of the proc
esses involved in the formation of these coastal sediments. It is seen that
the main factors controlling the composition of the clusters are the chemi
cal characteristics of the data and not the sources. A few exceptions are f
ound though, where direct point sources are identified. (C) 1999 Elsevier S
cience Ltd. All rights reserved.