PurposeThe evaluation of long-term heavy metal concentrations in the Gulf of Palermo (Italy) has been carried out in order to investigate how changes of pollution levels in the last 50 years can be reflected in marine sediments. Measurements of Cr, Cu, Hg, Pb, and Zn concentrations were performed on dated fractions of a sediment core. Time series analysis has allowed to obtain information on the chronology of the heavy metal pollution of the area and to identify seasonal components and trends.Materials and methodsHeavy metal concentrations in the <63 μm fraction of core sections were obtained by Atomic Absorption Spectrophotometry, after wet sieving, drying, and digestion procedures. Dating of sediment core had been previously achieved by excess 210Pb method. A methodological approach based on the time series analysis was used in order to investigate the presence of heavy metal cyclic variations and to outline the trend of each pollutant. A spectral analysis was performed on concentration data to investigate the presence of seasonal components. The additive component model, widely used to estimate seasonal and long-term behavior, was chosen for the time series analysis. R statistical software and advanced toolkits have been used.Results and discussionSediment core has been found to provide information for years from 1959 to 2004. Time series analysis results showed the presence of statistically significant cycles, of about 10.3, 7.2, and 4.8 years. Seasonal components of heavy metals concentrations and well known cycles have been compared. In particular, correlations with seasonal components of Sun Spot Number Cycle, Multivariate ENSO Index, and Winter North Atlantic Oscillation Index have been investigated. Moreover, the presence of trends in heavy metal concentration data has been highlighted.ConclusionsTime series analysis on the concentrations of heavy metals from a dated sediment core sampled in the Palermo Gulf (Italy) has been carried out. Cyclical pattern over a time of about 50 years have been identified thanks to spectral analysis. Statistically significant cycles have been detected, showing how metal concentrations in sediments also depend on natural cycles. Seasonal and trend components have been extracted, through time series decomposition by the LOESS procedures. Seasonal components of our data and of natural cycles show a good agreement in terms of series correlations. Sediment core, on a local scale and with good reliability, records the dynamics of heavy metal pollution of the area, also allowing comparisons on a much larger scale