Factor Analysis With Nonnegative Constraints Research Articles

Source characteristics of polycyclic aromatic hydrocarbons and polychlorinated biphenyls in surface soils of Shenyang, China: A comparison of two receptor models combined with Monte Carlo simulation

The surface soil concentrations of 16 PAHs and 15 PCBs were simultaneously determined by gas chromatography-tandem mass spectrometry in 21 locations of urban areas of Shenyang. The average concentrations of PAHs and PCBs were 26.40 ± 34.68 mg/kg and 48.03 ± 27.47 μg/kg, respectively. Factor analysis with nonnegative constraints (FA-NNC) and absolute principal component score with multiple linear regression (APCS-MLR) model were used to explore and evaluate the sources of PAHs and PCBs in the study area. The results of FA-NNC showed that PAHs in soils were mainly from traffic emissions (49.64%), coal combustion (46.88%) and petrogenic source (3.49%). The PCBs in soils were mainly from commercial and high temperature combustion mixed sources (20.3%), combustion and industry emission mixed sources (21.1%), electrical equipment sources (22.2%) and traffic emission sources (36.4%). The results of APCS-MLR were consistent with those of FA-NNC. The uncertainty of FA-NNC and APCS-MLR model was analyzed by Monte Carlo simulation method. The results revealed the reliability of the two receptor models on source apportionment. The estimated carcinogenic risks indicated that the risks of PAHs in soils exceed the acceptable range (10−6-10−4), while the risks of PCBs were below the acceptable risk level of 10−6.

An in-depth investigation of the influence of sample size on PCA-MLR, PMF, and FA-NNC source apportionment results.

The mechanism by which parameters influence the source apportionment results of receptor models is not well understood. Three mature receptor models, namely, principal component analysis-multiple linear regression (PCA-MLR), positive matrix factorization (PMF) and factor analysis with nonnegative constraints (FA-NNC), were comparatively employed for source apportionment of 16 polycyclic aromatic hydrocarbons in 30 street dust samples. The results indicated that the FA-NNC and PMF models produced results with a higher degree of similarity than the results obtained with the PCA-MLR model. Moreover, when the sample size was gradually decreased, similar source profiles were extracted that were consistent with results obtained from all samples. However, the overall contribution rates were not as stable as the source profiles. The PCA-MLR results remained the most stable in both aspects. FA-NNC and PMF performed better in regards to the stability of contribution rates and source profiles, respectively. Improvements in the goodness of fit of overall and individual pollutants were always accompanied by a decrease in the relevance among the variables, indicating that while the model simulation effect was improved, the credibility of the results decreased. Thus, finding an appropriate number of sample size is more appropriate than simply involving too many samples in source apportionment models.

Contamination and source apportionment of metals in urban road dust (Jinan, China) integrating the enrichment factor, receptor models (FA-NNC and PMF), local Moran's index, Pb isotopes and source-oriented health risk

Contamination and source identifications of metals in urban road dust are critical for remediation and health protection. Receptor models are commonly used for metal source identification, whereas the results are usually subjective and not verified by other indicators. Here we present and discuss a comprehensive approach to study metal contamination and sources in urban road dust (Jinan) in spring and winter by integrating the enrichment factor (EF), receptor models (positive matrix factorization (PMF) and factor analysis with nonnegative constraints (FA-NNC)), local Moran's index, traffic factors and Pb isotopes. Cadmium, Cr, Cu, Pb, Sb, Sn and Zn were the main contaminants, with mean EFs of 2.0–7.1. The EFs were 1.0–1.6 times higher in winter than in spring but exhibited similar spatial trends. Chromium contamination hotspots occurred in the northern area, with other metal contamination hotspots in the central, southeastern and eastern areas. The FA-NNC results indicated Cr contamination primarily resulting from industrial sources and other metal contamination primarily originating from traffic emissions during the two seasons. Coal burning emissions also contributed to Cd, Pb and Zn contamination in winter. FA-NNC model-identified metal sources were verified via traffic factors, atmospheric monitoring and Pb isotopes. The PMF model failed to differentiate Cr contamination from other detrital metals and the above anthropogenic sources, largely due to the model grouping metals by emphasizing hotspots. Considering the FA-NNC results, industrial and traffic sources accounted for 28.5 % (23.3 %) and 44.7 % (28.4 %), respectively, of the metal concentrations in spring (winter), and coal burning emissions contributed 34.3 % in winter. Industrial emissions primarily contributed to the health risks of metals due to the high Cr loading factor, but traffic emissions dominated metal contamination. Through Monte Carlo simulations, Cr had 4.8 % and 0.4 % possibilities posing noncarcinogenic and 18.8 % and 8.2 % possibilities posing carcinogenic risks for children in spring and winter, respectively.

Source Apportionment and Spatial Distribution Simulation of Heavy Metals in a Typical Petrochemical Industrial City

Identifying the quantitative source and hazardous areas of heavy metals in soils plays a pivotal role in soil pollution research, and can provide a basis for regional soil risk monitoring and environmental management. For this purpose, a total of 175 samples were collected in topsoils from Linzi, a typical petrochemical industrial city in Shandong Province. Positive matrix factorization (PMF) and factor analysis with non-negative constraints (FA-NNC) receptor models were applied to analyze the sources of the heavy metals. Based on the multivariate statistical simulation methods of min/max autocorrelation factors (MAF) and sequential Gaussian simulation (SGS), the distribution of heavy metal and potential pollution areas were determined. As, Co, Cr, and Mn were mainly affected by natural sources, their concentrations were dominated by the parent materials, and the high-value areas were distributed in the south of the study area. Hg was the most serious pollution element among the 10 heavy metals analyzed in Linzi and originated from atmosphere deposition from industrial emissions and coal combustion, and the highest values were distributed in the northeast of the study area. Cd, Cu, Ni, Pb, and Zn were dominated by natural sources and human activities. The hot-spot areas were mainly concentrated in the middle of the study area. The potentially contaminated areas of Cd and Hg were 580.80 km2 and 666.60 km2, about 85.04% and 97.59% of the total area, and should require more attention. The potential pollution area of most elements was small and scattered across the study area, accounting for less than 1%.

Source apportionment and spatial distribution of potentially toxic elements in soils: A new exploration on receptor and geostatistical models

Potentially toxic element (PTE) pollution is considered as the main soil environmental problem in the world. Source apportionment and spatial pattern of soil PTEs are essential for soil management. US-EPA positive matrix factorization (EPAPMF) and sequential Gaussian simulation (SGS) are general modeling tools for source apportionment and spatial distribution, respectively. Factor analysis with nonnegative constraints (FA-NNC) and stochastic partial derivative equations (SPDE) provided potential tools for this issue. We compared the performance of FA-NNC with PMF and the performance of SPDE with SGS, based on a dataset containing 9 PTEs in 285 topsoil samples. Three factors were determined by the two receptor models, and the source contributions were similar, suggesting that FA-NNC can validly identify quantitative sources of soil PTEs. The average source contributions were calculated based on the PMF and FA-NNC. Natural sources dominated the contents of As, Co, Cr, Cu, Ni, and Zn and affected 56.0%, 38.7%, and 84.8% of the Cd, Hg, and Pb concentrations, respectively. A total of 59.8% of Hg and 12.0% of Pb were associated with atmospheric deposition from coal combustion, industrial and traffic emissions, respectively. Agricultural and industrial activities contributed 37.2% of Cd concentration. SPDE proved to be an effective geostatistical technique to simulate the spatial patterns of soil PTEs with higher prediction accuracy than SGS. Co, Cr, Cu, and Ni had similar spatial patterns with hotspots randomly distributed across the study area. The common hotspots of As, Cd, Hg, Pb, and Zn in central parts inherited their high geochemical background in mudstone, while intensive human inputs in these areas also contributed to the accumulation of Cd, Hg, and Pb.

Profile and source apportionment of volatile organic compounds from a complex industrial park.

Industrial emissions, mainly from industrial parks, are important sources of ambient volatile organic compounds (VOCs). Identification of the major sources of VOCs from industrial parks has practical significance in emission reduction. In this study, the major species of VOCs from a residential area located downwind of a complex industrial park were sampled with Tenax absorption tubes and analyzed by thermal desorption coupled with gas chromatography/mass spectrometry (TD-GC/MS). Receptor models of factor analysis with nonnegative constraints (FA-NNC) and positive matrix factorization (PMF) were employed to recognize the potential emission sources, which suggested an association with the production processes in the nearby industrial park. In order to validate the sources, the profiles of VOC emissions of related workshops under actual manufacturing processes were acquired. It was found that xylenes & amines, phenols and esters were the major species of VOCs for the workshops of foundry, refractory materials and printing, respectively. Similarity analysis indicated that the detected profiles of VOC emissions from the dominant industrial types had good correlations with the identified factors from receptor models. Source contributions to VOCs in the receptor region exhibited that foundry production was the primary contributor (56-64%), followed by refractory material production (22-26%) and printing (14-18%). This study provides a strategy for source apportionment of VOCs from a local complex industrial park, which is helpful in the development of targeted control strategies.

Source Apportionment of Polycyclic Aromatic Hydrocarbons in Sediment by the Application of Non-Negative Factor Analysis: A Case Study of Dalian Bay.

An improved method, factor analysis with non-negative constraints (FA-NNC) was adopted to apportion the sources of sediment polycyclic aromatic hydrocarbons (PAHs) in Dalian Bay, China. Cosine similarity and Monte Carlo uncertainty analysis were used to assist the FA-NNC source resolution. The results identified three sources for PAHs, which were overall traffic, diesel engine emissions and residential coal combustion. The contributions of these sources were quantified as 78 ± 4.6% from overall traffic, 12 ± 3.2% from diesel engine emissions, and 10 ± 1.9% from residential coal combustion. The results from the Monte Carlo uncertainty analysis indicated that the model was robust and convergent.

Source apportionment of trace metals in river sediments: A comparison of three methods

Increasing trace metal pollution in river sediment poses a significant threat to watershed ecosystem health. Identifying potential sources of sediment metals and apportioning their contributions are of key importance for proposing prevention and control strategies of river pollution. In this study, three advanced multivariate receptor models, factor analysis with nonnegative constraints (FA-NNC), positive matrix factorization (PMF), and multivariate curve resolution weighted-alternating least-squares (MCR-WALS), were comparatively employed for source apportionment of trace metals in river sediments and applied to the Le'an River, a main tributary of Poyang Lake which is the largest freshwater lake in China. The pollution assessment with contamination factor and geoaccumulation index suggested that the river sediments in Le'an River were contaminated severely by trace metals due to human activities. With the three apportionment tools, similar source profiles of trace metals in sediments were extracted. Especially, the MCR-WALS and PMF models produced essentially the same results. Comparatively speaking, the weighted schemes might give better solutions than the unweighted FA-NNC because the uncertainty information of environmental data was considered by PMF and MCR-WALS. Anthropogenic sources were apportioned as the most important pollution sources influencing the sediment metals in Le'an River with contributions of about 90%. Among them, copper tailings occupied the largest contribution (38.4–42.2%), followed by mining wastewater (29.0–33.5%), and agricultural activities (18.2–18.7%). To protect the ecosystem of Le'an River and Poyang Lake, special attention should be paid to the discharges of mining wastewater and the leachates of copper tailing ponds in that region.

Source identification of polycyclic aromatic hydrocarbons in different ecological wetland components of the Qinkenpao Wetland in Northeast China

To determine the distribution and sources of polycyclic aromatic hydrocarbon (PAH), sixteen common PAHs were measured in various wetland components, including groundwater, surface water, sediments and soils collected from the Qinkenpao Wetland of Daqing City in Northeast China. High levels of PAHs were observed in the wetland due to the discharge of industrial wastewater including petrochemical products, such as petroleum and carbolic acid, which can be traced back for nearly three decades. Diagnostic ratios of selected PAH compounds showed that PAHs were generated by a number of mixed sources dominated by petrogenic products. Factor analysis with nonnegative constraints (FA-NNC) was combined with Latin Hypercube Sampling (LHS) simulation, which is effective for using a relatively small sample size while preserving the desirable probabilistic features of simple random sampling, to quantitatively identify sources, source contribution, and uncertainty of PAH contamination. The profiles derived from FA-NNC coupled with LHS were compared with source fingerprints, which were modified based on the first order degradation reaction in different wetland components. Premium gasoline (high-octane) was found to be the largest contributor of PAHs in surface water, groundwater, and sediments (96.0 percent, 61.6 percent and 53.1 percent, correspondingly), while regular gasoline was the dominant source of PAHs in soils (49.0 percent). The results were in agreement with the area׳s environmental data and diagnostic ratio results. Diagnostic ratios and FA-NNC were shown to be an effective methodology for source identification of PAHs in different wetland components.

Source Apportionment of Water Pollution in the Jinjiang River (China) Using Factor Analysis With Nonnegative Constraints and Support Vector Machines

Source apportionment studies of water pollution can greatly improve the knowledge of the human impact on the aquatic environment. Factor analysis (FA) has been widely used to identify sources of water pollution because of its relative ease of implementation. Generally, the method of identifying the sources was by qualitatively comparing source emission characteristics with factor loadings derived from FA. However, this traditional method was somewhat coarse to express the nonlinear relationship between source emission characteristics and factor loadings. In this study, by treating source identification using source emission characteristics and factor loadings as a pattern recognition problem, a source apportionment method was proposed by combining the factor analysis with nonnegative constraints (FA-NNC) with the support vector machine (SVM). Data sets on water quality of the Jinjiang River (China), which were sampled between May 2009 and September 2010 at 13 sites, have been collected to evaluate this proposed method. The apportionment results showed that the identified sources using the combined models were similar to the comprehensive analysis results obtained from qualitatively comparing source emission characteristics with factor loadings. Industrial activities, including papermaking and textiles, metal handicrafts manufacture, chemical and metal producing, metal refining and iron ore mining were identified as the main pollution sources with contribution ratio of 79.58%, followed by agricultural non-point sources (20.42%). These results provide policy and decision makers with a useful help for supporting the management of water pollution in the Jinjiang River. Meanwhile, this study will provide a useful direction for developing source apportionment approach to support the management of water pollution.

A framework of characteristics identification and source apportionment of water pollution in a river: a case study in the Jinjiang River, China

A framework for characteristics identification and source apportionment of water pollution in the Jinjiang River of China was proposed in this study for evaluation. A total of 114 water samples which were generated between May 2009 and September 2010 at 13 sites were collected and analysed. First, support vector machine (SVM) and water quality pollutant index (WQPI) were used for water quality comprehensive evaluation and identifying characteristic contaminants. Later, factor analysis with nonnegative constraints (FA-NNC) was employed for source apportionment. Finally, multi-linear regression of the absolute principal component score (APCS/MLR) was applied to further estimate source contributions for each characteristic contaminant. The results indicated that the water quality of the Jinjiang River was mainly at the third level (65.79%) based on national surface water quality permissible standards in China. Ammonia nitrogen, total phosphorus, mercury, iron and manganese were identified as characteristic contaminants. Source apportionment results showed that industrial activities (63.16%), agricultural non-point source (16.50%) and domestic sewage (12.85%) were the main anthropogenic pollution sources which were influencing the water quality of Jinjiang River. This proposed method provided a helpful framework for conducting water pollution management in aquatic environment.

Source apportionment of polycyclic aromatic hydrocarbons (PAHs) in surface sediments of the Rizhao coastal area (China) using diagnostic ratios and factor analysis with nonnegative constraints

In this study, sources of polycyclic aromatic hydrocarbons (PAHs) found in surface sediments of the Rizhao coastal area (China) were apportioned using diagnostic ratios and factor analysis with nonnegative constraints (FA-NNC). Bivariate plots of selected diagnostic ratios showed that the sources of PAHs identified in surface sediments seemed to be mixed sources dominated by petroleum-related. Literature PAH source profiles were modified based on the first-order degradation reaction in the atmosphere and sediments, and were considered as comparison for source identification. Five significant factors were determined with the diagnostic tools including coefficient of determination, cumulative percent variance and Exner function. By visually comparing PAH patterns and from the sum of squares of differences between modeled and modified literature PAH profiles, the potential sources were apportioned with the FA-NNC. The main contribution sources of PAHs originated from diesel engine (27.22%), followed by traffic emission (25.03%), gasoline engine (18.95%), coal power plant (14.77%) and coal residential (14.03%). Energy consumption was the predominant reason for PAH pollution in that region.

A comparative study on source apportionment of polycyclic aromatic hydrocarbons in sediments of the Daliao River, China: Positive matrix factorization and factor analysis with non-negative constraints

Two receptor models, positive matrix factorization (PMF) and factor analysis with non-negative constraints (FA-NNC), were applied for source apportionment of polycyclic aromatic hydrocarbons (PAHs) in sediments of the Daliao River, China. The source profiles and source contributions derived from the two receptor models were comparable. For PMF, PAHs mainly came from coal combustion (43.3%), followed by biomass burning (24.3%), traffic emissions (16.7%) and coke oven (15.7%). As for FA-NNC, coal combustion (50.4%) was dominant, followed by biomass burning (21.9%), traffic emission (15.7%) and coke oven (12.0%). The source apportionment results are consistent with the results from the specific PAH diagnostic ratios and emission inventory analysis. In addition, the source apportionment results are consistent with the structure of energy material usage in Liaoning Province of China.

Application of Factor Analysis with Nonnegative Constraints for Source Apportionment of Soil Polycyclic Aromatic Hydrocarbons (PAHs) in Liaoning, China

An improved method, factor analysis with nonnegative constraints (FA-NNC) was employed to determine the source contributions of soil polycyclic aromatic hydrocarbons (PAHs) in Liaoning province, China, based on the measured PAH concentrations of 45 surface soil samples. The summation of 14 PAHs (Σ PAHs) ranged from 125 to 4115 ng/g dry weight with a median value of 533 ng/g. The high proportion of phenanthrene (Phe) (18.3%) implied that PAHs mainly originated from coal and/or biomass combustion. According to the results obtained from the FA-NNC, four sources were identified representing coal combustion, coke oven, biomass burning, and a labeled “other” source that may be a mixture profile. The contributions of these sources were quantified as 50.5% from coal combustion, 27.0% from coke oven, 19.4% from biomass burning and 3.1% from the “other” source. The results were compatible with emission estimation based on annual fuel consumption in the study area and emission factors of PAHs.

正定矩阵因子分解和非负约束因子分析用于大辽河沉积物中多环芳烃源解析的比较研究

Two receptor models, positive matrix factorization (PMF) and factor analysis with non-negative constraints (FA-NNC), were applied for source apportionment of polycyclic aromatic hydrocarbons (PAHs) in sediments of the Daliao River, China. The source profiles and source contributions derived from the two receptor models were comparable. For PMF, PAHs mainly came from coal combustion (43.3%), followed by biomass burning (24.3%), traffic emissions (16.7%) and coke oven (15.7%). As for FA-NNC, coal combustion (50.4%) was dominant, followed by biomass burning (21.9%), traffic emission (15.7%) and coke oven (12.0%). The source apportionment results are consistent with the results from the specific PAH diagnostic ratios and emission inventory analysis. In addition, the source apportionment results are consistent with the structure of energy material usage in Liaoning province of China.