Machine learning models for aerosol particle size hygroscopic growth factor
MI Jia-yuan1, LI Na1, TONG Jing-zhe2, NI Chang-jian1
1. College of Atmospheric Science, Chengdu University of Information Technology, Chengdu 610225, China; 2. Liaoning Provincial Meteorological Equipment Support Center, Shenyang 110166, China
Abstract:Based on the hourly observational data of nephelometer, aethalometer and GRIMM180 environment particle monitor from October to December 2017 in Chengdu, as well as the simultaneous data of atmospheric visibility (V), relative humidity (RH) and nitrogen dioxide (NO2), aerosol hygroscopic growth factor (Gf) was retrieved by the aid of Mie scattering theory and immune evolutionary algorithm. Firstly, RH, CBC, CBC/CPM1, CPM1/CPM2.5 and CPM2.5/CPM10 were used as explanatory variables set, three machine learning models for aerosol particle size hygroscopic growth factors were constructed (XGBoost model, CatBoost model, and LightGBM model), and the corresponding judgment coefficients (R2) were 0.869, 0.893 and 0.898, root mean square error (RMSE) were 0.108, 0.097 and 0.090, mean absolute error (MAE) were 0.061, 0.054 and 0.052, respectively. Tests of three models showed that, machine learning models for aerosol particle size hygroscopic growth significantly reduced the simulation bias of traditional univariate aerosol particle size hygroscopic growth models under high humidity conditions, and it also improved the calculation accuracy of multivariate GAM model for aerosol particle size hygroscopic growth. Finally, the effects on different explanatory variables of the simulation results of machine learning models were analyzed, black carbon was confirmed as the main control variable in the aerosol hygroscopic growth model. The above study further explained the complexity of the multifactorial influences on aerosol particle size hygroscopic growth factors, and provided a new approach to scientifically characterisation of Gf models.
韩永,王体健,饶瑞中,等.大气气溶胶物理光学特性研究进展[J]. 物理学报, 2008,11:7396-7407. Han Y, Wang T J, Rao R Z, et al. Progress in the study of physic-optics characteristics of atmospheric aerosols [J]. Acta Physica Sinica, 2008,11:7396-7407.
[2]
Bamford H A, Baker J E. Nitro-polycyclic aromatic hydrocarbon concentrations and sources in urban and suburban atmospheres of the Mid-Atlantic region [J]. Atmospheric Environment, 2003,37:2077-2091.
[3]
Pope C A, Dockery D W. Heath effects of fine particulate air pollution:lines that connect [J]. Air & Waste Management Association, 2006, 56:709-742.
[4]
Hand J L, Kreidenweis S M, Sherman D E, et al. Aerosol size distributions and visibility estimates during the Big Bend regional aerosol and visibility observational (BRAVO) study [J]. Atmospheric Environment, 2002,36:5043-5055.
[5]
刘新罡,张远航.大气气溶胶吸湿性质国内外研究进展[J]. 气候与环境研究, 2010,15(6):808-816. Liu X G, Zhang Y H. Advances in research on aerosol hygroscopic properties at home and abroad [J]. Climatic and Environmental Research, 2010,15(6):808-816.
[6]
Yan P, Pan X, Jie T, et al. Hygroscopic growth of aerosol scattering coefficient:A comparative analysis between urban and suburban sites at winter in Beijing [J]. Particuology, 2009,7(1):52-60.
[7]
庄雯雯.基于地面MPL研究上海霾天气及吸湿增长的影响[D]. 上海:东华大学, 2015. Zhuang W W. The ground of micro pulse lidar based on strength, Shanghai haze analysis and moisture absorption characteristic of season [D]. Shanghai:Donghua University, 2015.
[8]
张智察,倪长健,邓也,等.免疫进化算法反演均匀混合气溶胶吸湿增长因子[J]. 中国环境科学, 2020,40(3):1008-1015. Zhang Z C, Ni C J, Deng Y, et al. Retrieval of hygroscopic growth factor of uniformly mixed aerosol particles based on immune evolution algorithm [J]. China Environmental Science, 2020,40(3):1008-1015.
[9]
Kasten F. Visibility forecast in the phase of pre-condensation [J]. Tellus, 1969,21(5):631-635.
[10]
孙景群.湿气溶胶的光散射特性[J]. 高原气象, 1983,2(3):49-54. Sun J Q. Relationship between visibility and relative humidity [J]. Plateau Meteorology, 1983,2(3):49-54.
[11]
张智察,倪长健,张城语,等.气溶胶粒径吸湿增长与散射吸湿增长的关系[J]. 中国环境科学, 2020,40(12):5198-5204. Zhang Z C, Ni C J, Zhang C Y, et al. Relationship between particle size hygroscopic growth and scattering hygroscopic growth [J]. China Environmental Science, 2020,40(12):5198-5204.
[12]
Zhang D, Zhang R Y. Laboratory investigation of heterogeneous interaction of sulfuric acid with soot [J]. Environmental Science & Technology, 2005,39:5722-5728.
[13]
谭天怡,郭松,吴志军,等.老化过程对大气黑碳颗粒物性质及其气候效应的影响[J]. 科学通报, 2020,65(36):4235-4250. Tan T Y, Guo S, Wu Z J, et al. Impact of aging process on atmospheric black carbon aerosol properties and climate effects [J]. Chin Sci Bull, 2020,65(36):4235-4250.
[14]
张城语,倪长健,佟景哲,等.成都地区气溶胶散射吸湿增长因子双变量模型[J]. 中国环境科学, 2021,41(12):5467-5475. Zhang C Y, Ni C J, Tong J Z, et al. Bivariate model of aerosol scattering hygroscopic growth factor in Chengdu [J]. China Environmental Science, 2021,41(12):5467-5475.
[15]
张智察,倪长健,汤津赢,等."干"气溶胶等效复折射率与其质量浓度指标的相关性研究[J]. 光学学报, 2019,39(5):17-24. Zhang Z C, Ni C J, Tang J Y, et al. Correlation between equivalent complex refraction index of "dry" aerosol and its mass concentration index [J]. Acta Optica Sinica, 2019,39(5):17-24.
[16]
佟景哲,倪长健,杜云松,等.成都气溶胶散射吸湿增长因子多变量影响分析[J]. 生态与农村环境学报, 2022,38(5):636-644. Tong J Z, Ni C J, Du Y S, et al. lmpact Analysis of Multivariable of Aerosol Scattering Hygroscopic Growth Factor in Chengdu [J]. Journal of Ecology and Rural Environment, 2022,38(5):636-644.
[17]
Chen J, Zhao C S, Ma N, et al. A parameterization of low visibilities for hazy days in the North China Plain [J]. Atmospheric Chemistry and Physics, 2012,12(271).
[18]
米家媛,佟景哲,倪长健,等.气溶胶粒径吸湿增长多因素影响的GAM模型[J]. 环境科学与技术, 2023,46(6):128-135. Mi J Y, Tong J Z, Ni C J, et al. GAM Model for Multifactorial Effect on Aerosol Particle Size Hygroscopic Growth [J]. Environmental Science & Technology, 2023,46(6):128-135.
[19]
王孟,覃露,王春娟,等.基于机器学习算法的脑出血相关肺炎预测模型研究[J]. 中国卒中杂志, 2020,15(3):243-249. Wang M, Qin L, Wang C J, et al. Machine Learning-based Models for Prediction of Intracerebral Hemorrhage Associated Pneumonia [J]. Chinese Journal of Stroke, 2020,15(3):243-249.
[20]
唐嘉豪.基于机器学习的航班延误预测方法研究[D]. 广汉:中国民用航空飞行学院, 2021. Tang J H. Research on Flight Delays Prediction Methods Based on Machine Learning [D]. Guanghan:Civil Aviation Flight University of China, 2021.
[21]
胡占占,陈传法,胡保健.基于时空XGBoost的中国区域PM2.5浓度遥感反演[J]. 环境科学学报, 2021,41(10):4228-4237. Hu Z Z, Chen C F, Hu B J. Estimating PM2.5 concentrations across China based on space-time XGBoost approach [J]. Acta Scientiae Circumstantiae, 2021,41(10):4228-4237.
[22]
Ke G, Meng Q, Finley T, et al. LightGBM:A highly efficient gradient boosting decision tree [J]. Advances in neural information processing systems, Long Beach, CA, USA:NIPS Foundation, 2017,30:3149-3157.
[23]
Dorogush A, Ershov V, Gulin A. CatBoost:Gradient boosting with calegorical fealures support [C]. Neural Information Processing Systems. Long Beach, CA, USA:NIPS Foundation, 2017:1-7.
[24]
Chen T, Guestrin C. XGBoost:a scalable tree boosting system [C]. ACM SIGKDD international conference on knowledge discovery and data mining. ACM, 2016:785-794.
[25]
Weingartner E, Burtscher H, Baltensperger U. Hygroscopic properties of carbon and diesel soot particles [J]. Atmospheric Environment, 1997,31(15):2311-2327.
[26]
Posfai M, Anderson J R, Buseck P R, et al. Soot and sulfate aerosol particles in the remote marine troposphere [J]. Journal of Geophysical Research Atmospheres, 1999,104(D17):21685-21693.
[27]
刘凡,谭钦文,江霞,等.成都市冬季相对湿度对颗粒物浓度和大气能见度的影响[J]. 环境科学, 2018,39(4):1466-1472. LIU F, TAN Q W, JIANG X, et al. Effect of relative humidity on particulate matter concentration and visibility during winter in Chengdu [J]. Environmental Science, 2018,39(4):1466-1472.
[28]
Cappa C D, Zhang X L, Russell L M, et al. Light absorption by ambient black and brown carbon and its dependence on black carbon coating state for two California, USA, cities in winter and summer [J]. Journal of Geophysical Research:Atmospheres, 2019,124:1550−157.
[29]
Zhang F, Peng J F, Chen L, et al. The effect of Black carbon aging from NO2 oxidation of SO2 on its morphology, optical and hygroscopic properties [J]. Environmental Research, 2022,212:113238.
[30]
姚婷婷,黄晓锋,何凌燕,等.深圳市冬季大气消光性质与细粒子化学组成的高时间分辨率观测和统计关系研究[J]. 中国科学:化学, 2010,40(8):1163-1171. Yao T T, Huang X F, He L Y, et al. High time resolution observation and statistical analysis of atmospheric light extinction properties and the chemical speciation of fine particulates [J]. Scientia Sinica (Chimica), 2010,40(8):1163-1171.