基于梯度提升回归树算法的地面臭氧浓度估算

摘要
图/表
参考文献(49)
相关文章 (15)

全文: PDF (5069 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要

将机器学习中的梯度提升回归树（GBRT）算法应用到中国地区地面O₃浓度制图中，利用地面O₃浓度观测数据，结合WRF气象数据、MODIS植被归一化指数以及高程人口数据建立训练预测数据集.通过反向变量选择法选取模型最佳特征变量对其进行训练，十折交叉验证结果：决定系数R²=0.89、均方根误差RMSE=4.75μg/m³.同时对全国O₃人口暴露水平进行评估.结果表明：在暴露强度上，我国人口加权O₃浓度值排在前5的省依次是山东、河南、江苏、河北、上海，均值浓度为94.48μg/m³.在暴露持续时间上，非达标天数最多的5个省依次是河南、山东、河北、宁夏、北京，一年内有42%的天数处于非达标的状态.

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	李一蜚
	秦凯
	李丁
	樊文智
	何秦

关键词 ：臭氧(O₃), 梯度提升回归树(GBRT), 人口暴露, 时空分布

Abstract：

Gradient Boosting Regression Tree (GBRT) algorithm has been developed for application to the mapping of ground-based O₃ concentrations in China. In this study, observations of ground-based O₃ concentrations, WRF meteorological data, MODIS vegetation normalization index, elevation and population data were used to build a training prediction data set. The best feature variable of the model was selected using an inverse variable selection method. The results were evaluated using a ten-fold cross-validation method:the coefficient of determination R²=0.89and the root mean square error RMSE=4.75μg/m³. The model was used to evaluate the national O₃ population exposure level. In terms of exposure intensity, the provinces in China with the highest population-weighted O₃ concentration values are Shandong, Henan, Jiangsu, Hebei, and Shanghai, with an average concentration of 94.48μg/m³. In terms of exposure duration, the provinces with the highest number of non-compliant days are Henan, Shandong, Hebei, Ningxia, and Beijing, with an average percentage of 42% of thedays per year which are non-compliant.

Key words： O₃ Gradient Boosting Regression Tree (GBRT) population exposure temporal and spatial distribution

收稿日期: 2019-08-09

PACS:

X511

基金资助:

国家自然科学基金项目（41975041）；徐州重点研发计划（社会发展）（KC18225）

通讯作者: 秦凯,副教授,qinkai@cumt.edu.cn E-mail: qinkai@cumt.edu.cn

作者简介: 李一蜚(1993-),男,甘肃靖远人,中国矿业大学环境与测绘学院硕士研究生,主要从事大气环境遥感研究.发表论文3篇.

引用本文:

李一蜚, 秦凯, 李丁, 樊文智, 何秦. 基于梯度提升回归树算法的地面臭氧浓度估算[J]. 中国环境科学, 2020, 40(3): 997-1007. LI Yi-fei, QIN Kai, LI Ding, FAN Wen-zhi, HE Qin. Estimation of ground-level ozone concentration based on GBRT. CHINA ENVIRONMENTAL SCIENCECE, 2020, 40(3): 997-1007.

链接本文:

http://www.zghjkx.com.cn/CN/ 或 http://www.zghjkx.com.cn/CN/Y2020/V40/I3/997

[1]	Ii J E S, Suggs J C, Vorburger L M. Ozone indicators determined at rural sites in the eastern united states by two monitoring networks[J]. Environmental Monitoring & Assessment, 2004,65(3):485-502.
[2]	廖宏,漏嗣佳,朱彬.中国地区臭氧前体物对地面臭氧的影响[J]. 大气科学学报, 2010,33(4):451-459. Miu h, Lou S J, Zhu B. Impacts of O3 precursor on surface O3 concentration over China[J]. Transactions of Atmospheric Sciences, 2010,33(4):451-459.
[3]	Krotkov N A, Mclinden C A, Li C, et al. Aura OMI observations of regional SO2 and NO2 pollution changes from 2005 to 2014[J]. Atmospheric Chemistry & Physics Discussions, 2015,15(19):26555-26607.
[4]	刘建,吴兑,范绍佳,等.前体物与气象因子对珠江三角洲臭氧污染的影响[J]. 中国环境科学, 2017,37(3):813-820. Liu J, Wu D, Fan S J, et al. Impacts of precursors and meteorological factors on ozone pollution in Pearl River Delta[J]. China Environmental Science, 2017,37(3):813-820.
[5]	刘启明,方月敏,黄志勇,等.大气臭氧污染的生物学指标监测评价[J]. 生态环境学报, 2011,20(4):612-615. Liu Q M, Fang Y M, Huang Z Y, et al. Biological indicators for air quality assessment effects of ozone[J]. Ecology and Environment, 2011,20(4):612-615.
[6]	Bell M L. The use of ambient air quality modeling to estimate individual and population exposure for human health research:A case study of ozone in the Northern Georgia Region of the United States[J]. Environment International, 2006,32(5):586-593.
[7]	Guo Y, Zeng H, Zheng R, et al. The association between lung cancer incidence and ambient air pollution in China:A spatiotemporal analysis[J]. Environmental Research, 2016,144(Pt A):60-65.
[8]	Shang Y, Sun Z, Cao J, et al. Systematic review of Chinese studies of short-term exposure to air pollution and daily mortality[J]. Environment International, 2013,54(4):100-111.
[9]	Sun Q, Wang W, Chen C, et al. Acute effect of multiple ozone metrics on mortality by season in 34 Chinese counties in 2013~2015[J]. Journal of Internal Medicine, 2017,283(5).
[10]	Adam-Poupart A, Brand A, Fournier M, et al. Spatiotemporal modeling of ozone levels in quebec (Canada):acomparison of kriging, Land-Use Regression (LUR), and combined bayesian maximum Entropy-LUR approaches[J]. Environmental Health Perspectives, 2014,122(9):970-976.
[11]	Madaniyazi L, Nagashima T, Guo Y, et al. Projecting ozone-related mortality in East China[J]. Environment International, 2016,92-93:165-172.
[12]	Zhan Y, Luo Y, Deng X, et al. Spatiotemporal prediction of daily ambient ozone levels across China using random forest for human exposure assessment[J]. Environmental Pollution, 2017,233:464.
[13]	丁愫,陈报章,王瑾,等.基于决策树的统计预报模型在臭氧浓度时空分布预测中的应用研究[J]. 环境科学学报, 2018,38(8):3229-3242. Ding X, Chen B Z, Wang J, et al. An applied research of decision-tree based statistical model in forecasting the spatial-temporal distribution of O3[J]. Acta Scientiae Circumstantiae, 2018,38(8):3229-3242.
[14]	美国环保局考虑建议8小时臭氧标准[J]. 中国环境科学, 1995, 15(4):271. U.S. Environmental Protection Agency considers suggesting 8-hour ozone standard[J]. China Environmental Science, 1995,15(4):271.
[15]	刘煜,周秀骥,李维亮.对流层臭氧的数值模拟试验[J]. 应用气象学报, 1990,1(1):45-56. Liu Y, Zhou X J, Li W L, et al. Numerical simulation of tropospheric ozone[J]. Journal of Applied Meteorological Science, 1990,1(1):45-56.
[16]	黄小刚,邵天杰,赵景波,等.气象因素和前体物对中国东部O3浓度分布的影响[J]. 中国环境科学, 2019,39(6):2273-2282. Huang X G, Shao T J, Zhao J B, et al. Impact of meteorological factors and precursors on spatial distribution of ozone concentration in Eastern China[J]. China Environmental Science, 2019,39(6):2273-2282.
[17]	郑中,祁元,潘小多,等.基于WRF模式数据和CASA模型的青海湖流域草地NPP估算研究[J]. 冰川冻土, 2013,35(2):465-474. Zhen Z, Qi Y, Pan X D, et al. Estimating the grassland NPP in Qinghai Lake Basin based on WRF model data and CASA model[J]. Journal of Glaciology and Geocryology, 2013,35(2):465-474.
[18]	吴雪娇,潘小多,沈永平,等.WRF模式制备的气象驱动数据在新疆喀依尔特斯河流域的验证[J]. 冰川冻土, 2016,38(2):332-340. Wu X J, Pan X D, Shen Y P, et al. Validation of meteorological driven data prepared by WRF model in Xinjiang Kailtes River Basin[J]. Journal of Glaciology and Geocryology, 2016,38(2):332-340.
[19]	朱求安,张万昌,余钧辉.基于GIS的空间插值方法研究[J]. 江西师范大学学报(自然科学版), 2004,28(2):183-188. Zhu Q A, Zhang W C, Yu J H. The spatial interpolations in GIS[J]. Journal of Jiangxi Normal Universiy (Natural Science Edition), 2004, 28(2):183-188.
[20]	岳文泽,徐建华,徐丽华.基于地统计方法的气候要素空间插值研究[J]. 高原气象, 2005,24(6):974-980. Yue W Z, Xu J H, Xu L H, et al. A Study on spatial interpolation methods for climate variables based on geostatistics[J]. Plateau Meteorology, 2005,24(6):974-980.
[21]	朱会义,贾绍凤.降雨信息空间插值的不确定性分析[J]. 地理科学进展, 2004,39(2):34-42. Zhu H Y, Jia S F. Uncertainty in the spatial interpolation of rainfall data[J]. Progress in Geography, 2004,39(2):34-42.
[22]	林忠辉,莫兴国,李宏轩,等.中国陆地区域气象要素的空间插值[J]. 地理学报, 2002,57(1):47-56. Lin Z H, Mo X G, Li H X, et al. Comparison of three spatial interpolation methods for climate variables in China[J]. Acta Geographica Sinica Acta Geogr. Sin., 2002,57(1):47-56.
[23]	王天送,张杰,孙明明.拉伊达准则在交通调查数据处理中的应用[J]. 西部交通科技, 2016,(4):96-99. Wang T S, Zhang J, Sun M M. Application of pautacriterion in traffic survey data processing[J].Western China Communications Science & Technology, 2016,(4):96-99.
[24]	李蕊,王慧忠,冯建文.改进拉伊达准则在LabVIEW温度采集系统中的应用[J]. 计算机测量与控制, 2018,26(6):207-210. Li R, Wang H S, Feng J W, et al. Application of pauta criterion in temperature acquisition system based on labVIEW[J]. Computer Measurement & Control, 2018,26(6):207-210.
[25]	Briggs D J, Collins S, Elliott P, et al. Mapping urban air pollution using GIS:a regression-based approach[J]. International Journal of Geographical Information Systems, 1997,11(7):699-718.
[26]	Hennig F, Sugiri D, Tzivian L, et al. Comparison of land-use regression modeling with dispersion and chemistry transport modeling to assign air pollution concentrations within the ruhr area[J]. Atmosphere, 2016,7(3):48.
[27]	Qin K, Zou J, Guo J, et al. Estimating PM1 concentrations from MODIS over Yangtze River Delta of China during 2014~2017[J]. Atmospheric Environment, doi:10.1016/j.atmosenv., 2018.09.054.
[28]	Yang B, Wu L, Kai Q, et al. A geographically and temporally weighted regression model for ground-level PM2.5 estimation from satellitederived 500m resolution AOD[J]. Remote Sensing, 2016,8(3):262.
[29]	Kai Q, Rao L, Jian X, et al. Estimating ground level NO2 concentrations over central-eastern china using a satellite-based geographically and temporally weighted regression model[J]. Remote Sensing, 2017,9(9):950.
[30]	赵佳楠,徐建华,卢德彬,等.基于RF-LUR模型的PM2.5空间分布模拟——以长江三角洲地区为例[J]. 地理与地理信息科学, 2018,34(1):18-23. Zhao J N, Xu J H, Lu D B, et al. The spatial distribution simulation of PM2.5 concentration based on RF-LUR model:Acase study of Yangtze River Delta[J]. Geography and Geo-information Science 2018,34(1):18-23.
[31]	刘小虎,李生.决策树的优化算法[J]. 软件学报, 1998,(10):78-81. Liu X H, Li S. Decision tree optimization algorithm[J]. Journal of Software, 1998,(10):78-81.
[32]	张宇.决策树分类及剪枝算法研究[D]. 哈尔滨:哈尔滨理工大学, 2009. Zhang Y. Research on decision tree classification and pruning algorithm[D]. Harbin:Harbin University of Science and Technology, 2009.
[33]	Mitchell T M, Carbonell J G, Michalski R S. Machine Learning[Z]. 1997.
[34]	游介文,邹滨,赵秀阁,等.基于随机森林模型的中国近地面NO2浓度估算[J]. 中国环境科学, 2019,39(3):969-979. You J W, Zhou B, Zhao X G, et al. Estimating ground-level NO2 concentrations across mainland China using random forests regression modeling[J]. China Environmental Science, 2019,39(3):969-979.
[35]	Friedman J H. Greedy function approximation:agradient boosting machine[J]. Annals of Statistics, 2001,29(5):1189-1232.
[36]	Schonlau M. Boosted Regression (Boosting):An introductory tutorial and a stata plugin[J]. Stata Journal, 2005,5(3):330-354.
[37]	Huang Y, Liu Y, Li C, et al. GBRTVis:online analysis of gradient boosting regression tree[J]. Journal of Visualization, 2019,22(1):125-140.
[38]	周鑫.基于GBRT的交通流量预测算法研究[J]. 现代计算机(专业版), 2019,(7):38-41. Zhou X. Research on the traffic flow forecasting algorithm based on GBRT[J]. Modern Computer (Professional Edition), 2019,(7):38-41.
[39]	李自法,谢维波,刘涛.一种基于GBRT算法的CA砂浆脱空检测方法[J]. 铁道科学与工程学报, 2018,15(2):292-301. Li Z F, Xie W B, Liu T. A detection method of CA mortar disengaging based on GBRT algorithm[J]. Journal of Railway Science and Engineering, 2018,15(2):292-301.
[40]	段晓瞳,曹念文,王潇,等.2015年中国近地面臭氧浓度特征分析[J]. 环境科学, 2017,38(12):4976-4982.Duan X T, Cao N W, Wang X, et al. Characteristics Analysis of the Surface Ozone Concentration of China in 2015[J]. Environmental Science, 2017,38(12):4976-4982.
[41]	郑永光,陈炯,陈尊裕,等.中国南部对流层中上层臭氧增加的气象场判识及臭氧变化的多尺度特征[J]. 地球物理学报, 2008,51(3):668-681. Zhen Y G, Chen X, Chen Z Y, et al. Identifying method of ozone enhancement and multi-scale characteristics of ozone change in the mid-upper troposphere over South China[J]. Chinese Journal of Geophysics, 2008,51(3):668-681.
[42]	Cardelino C A, Chameides W L. Natural hydrocarbons, urbanization, and urban ozone[J]. Journal of Geophysical Research Atmospheres, 1990,95(D9):13971-13979.
[43]	Chan L Y, Liu H Y, Lam K S, et al. Analysis of the seasonal behavior of tropospheric ozone at Hong Kong[J]. Atmospheric Environment, 1998,32(2):159-168.
[44]	Whelpdale D M, Moody J L. Large-scale meteorological regimes and transport process[Z]. 1990.
[45]	郭隽.通过WHo的空气质量准则看中国新颁布空气质量标准[J]. 资源节约与环保, 2013,(12):124. Guo X. Looking at China's newly issued air quality standards through WHO'sair quality guidelines[J]. Resources Economization & Environment Protection, 2013,(12):124.
[46]	任丽红,胡非,周德刚.城市边界层气象条件对O3浓度垂直分布的影响[J]. 城市环境与城市生态, 2004,(5):1-3. Ren L H, Hu F, Zhou D G. Effect of meteorological factors on O3 concentration vertical distribution in boundary layer[J]. Urban Environment & Urban EcologyCA, 2004,(5):1-3.
[47]	高文康,唐贵谦,辛金元,等.京津冀地区严重光化学污染时段O3的时空分布特征[J]. 环境科学研究, 2016,29(5):654-663. Gao W K, Tang G Q, Xin J Y, et al. Spatial-temporal variations of ozone during severe photochemical pollution over the BeijingTianjin-Hebei Region[J]. Research of Environmental Sciences, 2016, 29(5):654-663.
[48]	王男,王帅,刘闽,等.沈阳市O3时空分布特征及其与NOx的相关性研究[J]. 环境保护与循环经济, 2014,34(11):56-60. Wang N, Wang S, Liu M, et al. Study on the spatiotemporal distribution of O3 in Shenyang and its correlation with NOx[J]. Liaoning Urban and Rural Environmental Science & Technology, 2014,34(11):56-60.
[49]	苏筱倩,安俊琳,张玉欣.基于支持向量机回归和小波变换的O3预报方法[J]. 中国环境科学, 2019,39(9):3719-3726. Su Y Q, An J L, Zhang Y X. Support vector machine regression forecasting of O3 concentrations based on wavelet transformation[J]. China Environmental Science, 2019,39(9):3719-3726.