Pattern recognition on X-ray fluorescence records from Copenhagen lake sediments using principal component analysis

TY - JOUR

T1 - Pattern recognition on X-ray fluorescence records from Copenhagen lake sediments using principal component analysis

AU - Schreiber, Norman

AU - Garcia, Emanuel

AU - Kroon, Aart

AU - Ilsøe, Peter Carsten

AU - Kjær, Kurt H.

AU - Andersen, Thorbjørn Joest

PY - 2014/11/16

Y1 - 2014/11/16

N2 - Principal component analysis (PCA) was performed on chemical data of two sediment cores from an urban freshwater lake in Copenhagen, Denmark. X-ray fluorescence (XRF) core scanning provided the underlying datasets on 13 variables (Si, K, Ca, Ti, Cr, Mn, Fe, Ni, Cu, Zn, Rb, Cd, and Pb). Principal component analysis helped to trace geochemical patterns and temporal trends in lake sedimentation. The PCA models explained more than 80 % of the original variation in the datasets using only two or three principal components. The first principal component (PC1) was mostly associated with geogenic elements (Si, K, Fe, Rb) and characterized the content of minerogenic material in the sediment. In the case of both cores, PC2 was a good descriptor emphasized as the contamination component. It showed strong linkages with heavy metals (Cu, Zn, Pb), disclosing changing heavy-metal contamination trends across different depths. The sediments featured a temporal association with contaminant dominance. Lead contamination was superseded by zinc within the compound pattern which was linked to changing contamination sources over time. Principal component analysis was useful to visualize and interpret geochemical XRF data while being a straightforward method to extract contamination patterns in the data associated with temporal elemental trends in lake sediments.

AB - Principal component analysis (PCA) was performed on chemical data of two sediment cores from an urban freshwater lake in Copenhagen, Denmark. X-ray fluorescence (XRF) core scanning provided the underlying datasets on 13 variables (Si, K, Ca, Ti, Cr, Mn, Fe, Ni, Cu, Zn, Rb, Cd, and Pb). Principal component analysis helped to trace geochemical patterns and temporal trends in lake sedimentation. The PCA models explained more than 80 % of the original variation in the datasets using only two or three principal components. The first principal component (PC1) was mostly associated with geogenic elements (Si, K, Fe, Rb) and characterized the content of minerogenic material in the sediment. In the case of both cores, PC2 was a good descriptor emphasized as the contamination component. It showed strong linkages with heavy metals (Cu, Zn, Pb), disclosing changing heavy-metal contamination trends across different depths. The sediments featured a temporal association with contaminant dominance. Lead contamination was superseded by zinc within the compound pattern which was linked to changing contamination sources over time. Principal component analysis was useful to visualize and interpret geochemical XRF data while being a straightforward method to extract contamination patterns in the data associated with temporal elemental trends in lake sediments.

KW - Faculty of Science

KW - urban lake sediment

KW - contamination

KW - heavy metals

KW - XRF

KW - PCA

U2 - 10.1007/s11270-014-2221-5

DO - 10.1007/s11270-014-2221-5

M3 - Journal article

SN - 1567-7230

VL - 225

JO - Water, Air and Soil Pollution

JF - Water, Air and Soil Pollution

IS - 12

M1 - 2221

ER -