Simulation of heavy metal concentrations in PM2.5 based on nonlinear mathematics methods
LENG Xiang-zi1, WANG Qin-geng1,2, QIAN Xin1,2, LI Hui-ming1, LU Hao1
1. State Key Laboratory of Pollution Control and Resources Reuse, School of the Environment, Nanjing University, Nanjing 210023, China;
2. Jiangsu Collaborative Innovation Center of Atmospheric Environment and Equipment Technology, Nanjing University of Information Science & Technology, Nanjing 210044, China
Heavy metal concentrations were determined in PM2.5 samples collected from Gulou and Pukou campus of Nanjing University in 2013. The pollution characteristics of heavy metals in different districts and different seasons were analysed, respectively. The correlations between heavy metal concentrations with meteorological factors or conventional air pollutants were investigated, respectively. The obtained data were pre-processed by principal component analysis, followed by the establishment of rapid evaluation models based on Back Propagation Artificial Neural Network (BP-ANN) and Support Vector Machine (SVM), respectively. Then models obtained from these two nonlinear mathematics methods were compared with those from multiple linear regression model (MLR). The results showed that PM2.5 concentrations and the average heavy metal concentrations in PM2.5 were all highest in winner, followed by spring, whereas the lowest in summer and autumn. The significant linear correlations were found between heavy metal concentrations with meteorological factors or air pollutant concentrations. BP-ANN showed the highest correlation coefficients of training models for most tested heavy metals (except for Ba, Cr, and V). SVM showed the highest correlation coefficients of verified models for all the tested heavy metals. All three methods showed good modelling effects on the evaluation of Cd, Cu, Pb, Ni, and Zn, but relatively poor modelling effects on the evaluation of Cr, Fe, Sr, Ti, and V.
冷湘梓, 王勤耕, 钱新, 李慧明, 陆昊. 基于非线性数学方法的PM2.5中重金属浓度模拟[J]. 中国环境科学, 2017, 37(3): 821-828.
LENG Xiang-zi, WANG Qin-geng, QIAN Xin, LI Hui-ming, LU Hao. Simulation of heavy metal concentrations in PM2.5 based on nonlinear mathematics methods. CHINA ENVIRONMENTAL SCIENCECE, 2017, 37(3): 821-828.
Mohanraj R, Azeez P A, Priscilla T. Heavy Metals in Airborne Particulate Matter of Urban Coimbatore [J]. Archives of Environmental Contamination & Toxicology, 2004,47(2):162-167.
[2]
Bollati V, Marinelli B, Apostoli P, et al. Exposure to Metal-Rich Particulate Matter Modifies the Expression of Candidate MicroRNAs in Peripheral Blood Leukocytes [J]. Environmental Health Perspectives, 2010,118(6):763-768.
[3]
Ghio A J. Biological effects of Utah Valley ambient air particles in humans: a review [J]. Journal of Aerosol Medicine, 2004,17(2): 157-164.
[4]
Hornik K, Stinchcombe M, White H. Multilayer feedforward networks are universal approximators [J]. Neural Networks, 1989, 2(5):359-366.
[5]
瓦普尼克.统计学习理论的本质 [M]. 北京:清华大学出版社, 2000.
[6]
Sait C. Sofuoglu, Aysun Sofuoglu, Savas Birgili, et al. Forecasting Ambient Air SO2 Concentrations Using Artificial Neural Networks [J]. Energy Sources Part B Economics Planning & Policy, 2006,1(2):127-136.
[7]
Martín M L, Turias I J, González F J, et al. Prediction of CO maximum ground level concentrations in the Bay of Algeciras, Spain using artificial neural networks [J]. Chemosphere, 2008, 70(7):1190-5.
[8]
Hoffman S. Short-Time Forecasting of Atmospheric NOx Concentration by Neural Networks [J]. Environmental Engineering Science, 2006,23(4):603-609.
Mckendry I G. Evaluation of artificial neural networks for fine particulate pollution (PM10 and PM2.5) forecasting [J]. Journal of the Air & Waste Management Association, 2002,52(9):1096-101.
[11]
Kukkonen J, Partanen L, Karppinen A, et al. Extensive evaluation of neural network models for the prediction of NO2, and PM10, concentrations, compared with a deterministic modelling system and measurements in central Helsinki [J]. Atmospheric Environment, 2012,37(32):4539-4550.
[12]
Lu W Z, Wang W J. Potential assessment of the “support vector machine” method in forecasting ambient air pollutant trends [J]. Chemosphere, 2005,59(5):693-701.
[13]
Sánchez A S, Nieto P J G, Fernández P R, et al. Application of an SVM-based regression model to the air quality study at local scale in the Avilés urban area (Spain) [J]. Mathematical & Computer Modelling, 2011,54(5/6):1453-1466.
Chen J, Tan M, Li Y, et al. Characteristics of trace elements and lead isotope ratios in PM2.5 from four sites in Shanghai. [J]. Journal of Hazardous Materials, 2008,156(1-3):36-43.
Tao Y, Cheng X, Yang Z, et al. Distributional characteristics of heavy metal elements in atmospheric particulate matter and their impact on land quality in Liaoning province. [J]. Earth Science Frontiers, 2008.
Lin H P, Xiao-Yan W U, Zhan J M, et al. Analysis of atmospheric particles and heavy metals character at the area of Lanzhou City in summer and winter [J]. Zhongguo Huanjing Kexue/china Environmental Science, 2012,32(5):810-815.
[28]
Xu L, Chen X, Chen J, et al. Seasonal variations and chemical compositions of PM2.5, aerosol in the urban area of Fuzhou, China [J]. Atmospheric Research, 2012,104(1):264-272.
[29]
Duan J, Tan J, Wang S, et al. Size distributions and sources of elements in particulate matter at curbside,urban and rural sites in Beijing [J]. Journal of Environmental Sciences, 2012,24(1):87-94.
[30]
Fang G C, Wu Y S, Lee J F, et al. Characteristics and source identification study of ambient suspended particulates and ionic pollutants in an area abutting a highway [J]. Powder Technology, 2008,185(185):223-230.
[31]
Juhos I, Makra L, Tóth B. Forecasting of traffic origin NO and NO2, concentrations by Support Vector Machines and neural networks using Principal Component Analysis [J]. Simulation Modelling Practice & Theory, 2008,16(9):1488-1502.
[32]
Liao K J, Tagaris E, Napelenok S L, et al. Current and future linked responses of ozone and PM2.5 to emission controls. [J]. Environmental Science & Technology, 2008,42(13):4670-5.
[33]
Carpenter L J, Clemitshaw K C, Burgess R A, et al. Investigation and evaluation of the NOx/O3, photochemical steady state [J]. Atmospheric Environment, 1998,32(19):3353-3365.
[34]
陈 俏.支持向量机应用于大气污染物浓度预测 [D]. 西安:西安科技大学, 2010.
[35]
周 品. MATLAB神经网络设计与应用 [M]. 北京:清华大学出版社, 2013.
[36]
Feng Y, Zhang W, Sun D, et al. Ozone concentration forecast method based on genetic algorithm optimized back propagation neural networks and support vector machine data classification [J]. Atmospheric Environment, 2011,45(45):1979-1985.
[37]
Li H, Wang Q, Shao M, et al. Fractionation of airborne particulate-bound elements in haze-fog episode and associated health risks in a megacity of southeast China. [J]. Environmental Pollution, 2015,208:655-662.
