Study On Distribution And Combined Models Of Source Apportionment For PAHs In Liaohe Estuarine Reed Wetland

Due to the property of persistence, bioaccumulation, toxicity and long-range atmospheric transport, as well as carcinogenicity, mutagenicity and teratogenicity, persistent organic pollutants (POPs) have gained great attention through the world, and a number of studies on POPs have been conducted.16 EPA priority PAHs in soils from Liaohe reed wetland were measured using GC/MS. Distribution of PAHs in soils and Phragmites australis was expounded, and several influence factors were discussed. To make source apportionment of PAHs in soils, we applied three single receptor models, i.e., positive matrix factorization (PMF), Unmix and chemical mass balance (CMB) and constructed a combined model named Unmix-CMB to improve the accuracy of the apportionment result. Moreover, we combined the models with toxic equivalent quantity (TEQ) to quantitatively calculate the toxicity of different PAHs sources. It will be of great significance to control the pollution of PAHs in reed wetland soils.Range of PAHs concentration in soils was 235.3～373.8 ng/g, with a mean value of 287.0 ng/g, presenting a light pollution level. Concentration of soil PAHs was more easily influenced by total organic carbon in soils, and the clay content had also certain effect on total PAHs concentration, but the cation exchange capacity in soil influenced the high molecular weigh PAHs. On the distribution of PAHs in different water-stable aggregates, the content of total PAHs decreased with the decreasing grain size of aggregates, and the absorption ability to low molecular weight PAHs became stronger as the grain size of aggregates increased. We used L/H and diagnostic ratios to qualitatively infer that the PAHs sources in soils were mainly from petrogenic and combustion sources. Quantitative source apportionment was conducted by single and combined models. Result showed, PMF identified four main sources including petrogenic source (31%), gasoline engine emission (26%), diesel engine emission (23%) and biomass burning (20%); Unmix identified three factor including four sources (petrogenic source (43%), mixing sources of diesel and gasoline engine emission (35%), and biomass burning (22%)); Result of CMB showed, gasoline (29%) and diesel engine emission (28%) presented the highest contribution, followed by petrogenic source (22%) and biomass burning (21%); These pollution sources for PAHs probably had closed relation with human activities nearby such as traffic emission, burning lefted reeds, exploitation of petroleum, etc. Through the comparative study of three models, we can find that, all three models were performing well when analyzing the simulated and measured PAHs concentrations, CMB produced better fitness coefficient than PMF and Unmix, but the absolute values of error percentage for CMB were higher, and simulated values of PMF and Unmix model were close; We also find that CMB underestimated the contribution of PAHs source, and because of the collinearity problem, Unmix extracted the gasoline and diesel emission sources as a single factor; There are some differences between three models on the contribution of three common sources including petroleum sources, diesel and gasoline engine emission source, but little differences were found on biomass burning, which may have certain relationship with model selection algorithm and its parameters. To solve collinearity problem during apportionment process, we constructed Unmix-CMB model. Results of this combined model showed that the main sources of PAHs were petrogenic source (42.6%), gasoline engine emission (23.5%), biomass burning (22.3%) and diesel engine emission (11.7%). Results of toxicity source apportionment using PMF-TEQ, Unmix-TEQ, CMB-TEQ and Unmix-CMB-TEQ showed that, gasoline and diesel engine emission showed the highest toxicity contribution, with contribution rate of 64%,61%,96% and 63%, respectively. Contribution on toxicity from PMF-TEQ and Unmix-CMB-TEQ showed very similar results, and petrogenic source and biomass burning also showed a certain toxicity contribution in results from PMF-TEQ, Unmix-TEQ and Unmix-CMB-TEQ, but the toxicity contribution resulted from CMB-TEQ can be ignored, such difference between these results are closely related with the differences between models.