PDF(1281 KB)
PDF(1281 KB)
PDF(1281 KB)
PMF模型解析土壤重金属来源的不确定性
Uncertainty analysis of soil heavy metal source apportionment by PMF model
正定因子矩阵分解（PMF）是目前污染源解析领域应用最为广泛的受体模型之一，其不确定性研究一直是源解析研究的前沿和热点.利用拔靴法（BS）、替换法（DISP）和拔靴-替换法（BS-DISP）3种不确定性分析方法探讨了PMF模型应用于土壤重金属源解析的不确定性，并以德兴铜矿周边土壤重金属为对象开展案例研究.结果表明，6因子情景是PMF模型最佳运行结果；在6因子情景的源成分谱中，除Cr和Ti外，DISP和BS不确定性区间均处于标识元素基本值的0.6~1.5倍之间，BS-DISP不确定性区间处于基本值的0.6~1.6倍之间；模型结果的不确定性更多源于因子旋转误差.通过这3种不确定性分析方法可以获得PMF模型运算中的随机误差和因子旋转误差.其中，BS-DISP法和BS法得到的结果能够辅助判断因子数是否过拟合，并有助于理解源谱的不确定性，而DISP法能够用于理解旋转的不确定性，可作为评价旋转过程可行性的方法.
Positive matrix factorization (PMF) model is one of widely used technologies in pollutant source apportionment, and its uncertain analysis have always been the frontier issue as well as hotspot. Three error estimation methods, including bootstrap (BS), displacement (DISP) and bootstrap enhanced by displacement (BS-DISP), were used to evaluate the uncertainties of source apportionment by PMF model, and heavy metals in soils in Dexing, China were carried out as a case study. Six-factor scenario was the best solution for PMF model run, except for Cr and Ti, the uncertainty intervals of DISP and BS were between 0.6 and 1.5 times the basic value of the identified element and the BS-DISP uncertainty interval was between 0.6 and 1.6 times the basic value in the source profiles under six-factor scenario, the uncertainty of the model results was more due to the uncertainty generated in the factor decomposition process. The three uncertainty analysis methods could obtain the random error and factor rotation error in the operation of PMF model. Among them, the results obtained by BS-DISP and BS can assist in determining whether the factor number was over-fitting and help understand the uncertainty of the source profile. While, DISP could be used to understand the uncertainty of rotation and be used as a method to evaluate the feasibility of rotation process. This study provides a good sample for evaluating the reliability of soil heavy metal source apportionment that calculated by PMF model.
PMF模型 / 不确定性 / 土壤重金属 / 误差估计 / 源解析
error estimation / PMF model / soil heavy metal / source apportionment / uncertainty analysis
[1] 李娇,吴劲,蒋进元,等.近十年土壤污染物源解析研究综述[J]. 土壤通报, 2018,49(1):232-242. Li J, Wu J, Jiang J Y, et al. Review on source apportionment of soil pollutants in recent ten years[J]. Chinese Journal of Soil Science, 2018,49(1):232-242.
[2] 瞿明凯,李卫东,张传荣,等.基于受体模型和地统计学相结合的土壤镉污染源解析[J]. 中国环境科学, 2013,33(5):854-860. Qu M K, Li W D, Zhang C R, et al. Source apportionment of soil heavy metal Cd based on the combination of receptor model and geostatistics[J]. China Environmental Science, 2013,33(5):854-860.
[3] 张慧,郑志志,马鑫鹏,等.哈尔滨市土壤表层重金属污染特征及来源辨析[J]. 环境科学研究, 2017,30(10):1597-1606. Zhang H, Zheng Z Z, Ma X P, et al. Sources and pollution characteristics of heavy metals in surface soils of Harbin city[J]. Research of Environmental Sciences, 2017,30(10):1597-1606.
[4] Jiang Y, Chao S, Liu J, et al. Source apportionment and health risk assessment of heavy metals in soil for a township in Jiangsu Province, China[J]. Chemosphere, 2017,168:1658-1668.
[5] 艾建超,王宁,杨净.基于UNMIX模型的夹皮沟金矿区土壤重金属源解析[J]. 环境科学, 2014,35(9):3530-3536. Ai J C, Wang N, Yang J. Source apportionment of soil heavy metals in Jiapigou goldmine based on the UNMIX model[J]. Environmental Science, 2014,35(9):3530-3536.
[6] Wang G Q, A Y L, Jang H, et al. Modeling the source contribution of heavy metals in surficial sediment and analysis of their historical changes in the vertical sediments of a drinking water reservoir[J]. Journal of Hydrology, 2015,520:37-51.
[7] U.S. Environmental Protection Agency. EPA positive matrix factorization (PMF) 5.0 fundamentals and user guide[Z]. 2014.
[8] Guan Q, Wang F, Xu C, et al. Source apportionment of heavy metals in agricultural soil based on PMF:A case study in Hexi Corridor, northwest China[J]. Chemosphere, 2018,193:189-197.
[9] 董騄睿,胡文友,黄标,等.基于正定矩阵因子分析模型的城郊农田土壤重金属源解析[J]. 中国环境科学, 2015,35(7):2103-2111. Dong L R, Hu W Y, Huang B, et al. Source appointment of heavy metals in suburban farmland soils based on positive matrix factorization[J]. China Environmental Science, 2015,35(7):2103-2111.
[10] 李娇,陈海洋,滕彦国,等.拉林河流域土壤重金属污染特征及来源解析[J]. 农业工程学报, 2016,32(19):226-233. Li J, Chen H Y, Teng Y G, et al. Contamination characteristics and source apportionment of soil heavy metals in Lalin River basin[J]. Transactions of the Chinese Society of Agricultural Engineering, 2016,32(19):226-233.
[11] Khairy M A, Lohmann R. Source apportionment and risk assessment of polycyclic aromatic hydrocarbons in the atmospheric environment of Alexandria, Egypt[J]. Chemosphere, 2013,91(7):895-903.
[12] Yang B, Zhou L, Xue N, et al. Source apportionment of polycyclic aromatic hydrocarbons in soils of Huanghuai Plain, China:Comparison of three receptor models[J]. Science of The Total Environment, 2013,443:31-39.
[13] Zhang Y, Guo C S, Xu J, et al. Potential source contributions and risk assessment of PAHs in sediments from Taihu Lake, China:Comparison of three receptor models[J]. Water Research, 2012, 46(9):3065-3073.
[14] Lang Y H, Li G L, Wang X M, et al. Combination of Unmix and positive matrix factorization model identifying contributions to carcinogenicity and mutagenicity for polycyclic aromatic hydrocarbons sources in Liaohe delta reed wetland soils, China[J]. Chemosphere, 2015,120:431-437.
[15] Lang Y H, Li G L, Wang X M, et al. Combination of Unmix and PMF receptor model to apportion the potential sources and contributions of PAHs in wetland soils from Jiaozhou Bay, China[J]. Marine Pollution Bulletin, 2015,90(1/2):129-134.
[16] Song Y, Dai W, Shao M, et al. Comparison of receptor models for source apportionment of volatile organic compounds in Beijing, China[J]. Environmental Pollution, 2008,156(1):174-183.
[17] 田福林,陈景文,刘成雁,等.蒙特卡罗不确定性分析在受体模型来源解析中的应用[J]. 科学通报, 2011,56(32):2675-2680. Tian F L, Chen J W, Liu C Y, et al. Uncertainty analysis of source apportionment by Monte Carlo methods:A case study of factor analysis with non-negative constraints[J]. Chinese Science Bulletin, 2011,56(32):2675-2680.
[18] Keats A, Cheng M T, Yee E, et al. Bayesian treatment of a chemical mass balance receptor model with multiplicative error structure[J]. Atmospheric Environment, 2009,43(3):510-519.
[19] Larsen B R, Gilardoni S, Stenstrom K, et al. Sources for PM air pollution in the Po Plain, Italy:II. Probabilistic uncertainty characterization and sensitivity analysis of secondary and primary sources[J]. Atmospheric Environment, 2012,50(3):203-213.
[20] Paatero P, Eberly S, Brown S G, et al. Methods for estimating uncertainty in factor analytic solutions[J]. Atmospheric Measurement Techniques, 2014,7(3):781-797.
[21] Brown S G, Eberly S, Paatero P, et al. Methods for estimating uncertainty in PMF solutions:Examples with ambient air and water quality data and guidance on reporting PMF results[J]. Science of The Total Environment, 2015,518:626-635.
[22] Efron B, Tibshirani R. An introduction to the bootstrap[M]. Florida:CRC press, 1994.
[23] Davison A C, Hinkley D V. Bootstrap methods and their application[M]. London:Cambridge university press, 1997.
[24] Hall P. Theoretical Comparison of Bootstrap Confidence Intervals[J]. Annals of Statistics, 1988,16(3):927-953.
[25] 李娇,滕彦国,吴劲,等.基于PMF模型及地统计法的乐安河中上游地区土壤重金属来源解析[J]. 环境科学研究, 2019,32(6):984-992. Li J, Teng Y G, Wu J, Jiang J Y, Huang Y. Source apportionment of soil heavy metal in the middle and upper reaches of Le'an river based on PMF model and geostatistics[J]. Research of Environmental Sciences, 2019,32(6):984-992.
[26] Reff A, Eberly S I, Bhave P V. Receptor modeling of ambient particulate matter data using positive matrix factorization:review of existing methods[J]. Journal of the Air & Waste Management Association, 2007,57(2):146-154.
国家自然科学基金资助项目（41807344）；广西创新驱动发展专项资金资助项目（（AA17202032）
京公网安备 11010802030352号 /
| 〈 |
|
〉 |
