Improved source apportionment of PAHs and Pb by integrating Pb stable isotopes and positive matrix factorization application (PAHs): A historical record case study from the northern South China Sea

Abstract

To obtain the historical changes of pyrogenic sources, integrated source apportionment methods, which include PAH compositions, diagnostic ratios (DRs), Pb isotopic ratios, and positive matrix factorization (PMF) model, were developed and applied in sediments of the northern South China Sea. These methods provided a gradually clear picture of energy structural change. Spatially, Sigma(15)PAH (11.3 to 95.5 ng/g) and Pb (10.2 to 74.6 mu g/g) generally exhibited decreasing concentration gradient offshore; while the highest levels of PAHs and Pb were observed near the southern Taiwan Strait, which may be induced by accumulation of different fluvial input. Historical records of pollutants followed closely with the economic development of China, with fast growth of S15PAH and Pb occurring since the 1980s and 1990s, respectively. The phasing-out of leaded gasoline in China was captured with a sharp decrease of Pb after the mid-1990s. PAHs and Pb correlated well with TOC and clay content for core sediments, which was not observed for surface sediments. There was an up-core increase of high molecular PAH proportions. Coal and biomass burning were then qualitatively identified as the major sources of PAHs with DRs. Furthermore, shift toward less radiogenic signatures of Pb isotopic ratios after 1900 revealed the start and growing importance of industrial sources. Finally, a greater separation and quantification of various input was achieved by a three-factor PMF model, which made it clear that biomass burning, coal combustion, and vehicle emissions accounted for 40 +/- 20%, 41 +/- 13%, and 19 +/- 12% of PAHs through the core. Biomass and coal combustion acted as major sources before 2000, while contributions from vehicle emission soared thereafter. The integrated multi-methodologies here improved the source apportionment by reducing biases with a step-down and cross-validation perspective, which could be similarly applied to other aquatic systems. (C) 2017 Elsevier B.V. All rights reserved.