基于支持向量机和树叶磁学的大气颗粒物浓度模拟
|
摘要
在南京市仙林地区采集桂花、雪松及女贞树叶,测试其磁学参数,分析大气颗粒物浓度与树叶磁学参数、气象数据的相互关系,使用支持向量机方法构建大气颗粒物磁学评估模型。结果显示,雪松树叶的质量磁化率(χ)、饱和等温剩磁(SIRM)和非磁滞剩磁(ARM)值最高。3种树叶在夏季和秋季富集了相对较多的细磁畴磁性颗粒物,而在春季和冬季富集了相对较多的粗磁畴磁性颗粒物。PM_(10)和PM_(2.5)浓度与风速、温度和湿度之间显著负相关,与大气压显著正相关;与3种树叶的磁学参数显著正相关,且与χ和SIRM的相关性高于ARM,其中与桂花树叶χ、SIRM的相关系数均高于0.560。将"树叶磁学参数+气象数据"作为输入因子的支持向量机模型对PM_(2.5)浓度的模拟效果较好,将"树叶磁学参数"作为输入因子的模型对PM_(10)浓度的模拟效果较好,模型误差都在可接受范围内。
Samples of Osmanthus fragrans Lour, Ceder deodara G.Don and Ligustrum lucidum Ait were collected in Xianlin district of Nanjing. The magnetic parameters of tree leaf samples were measured. The correlations between concentrations of atmospheric particulate matter and magnetic parameter of leaves, as well as meteorological data were analyzed. The support vector machine(SVM) method was used to construct the magnetic evaluation model of atmospheric particulate matter. Results showed that the value of mass magnetization rate(χ), saturated isothermal remanence(SIRM) and anhysteretic remanent magnetization(ARM) of leaves of cedar were the highest. The three leaves accumulated relatively more fine-magnetic domain magnetic particles in summer and autumn, whereas relatively more coarse-magnetic domain magnetic particles in spring and winter. PM_(10) and PM_(2.5) concentrations showed significant negative correlation with wind speed, temperature, and humidity, and significant positive correlation with atmospheric pressure. The magnetic parameters of three species of leaves had positively correlations with PM_(10) and PM_(2.5) concentrations(p<0.05). The correlation coefficients(r) of magnetic parameters of tree leaves were higher with χ and SIRM than with ARM.The correlation coefficient(r) of magnetic parameters of Osmanthus fragrans Lour leaves with PM_(10) and PM_(2.5) concentrations were both higher than 0.560. Taking "leaf magnetic parameters and meteorological data" as the input factors, the simulation effect of SVM models on PM_(2.5) concentration was better, while the simulation effect of taking only "leaf magnetism parameter" as the input factor on PM_(10) concentration was better. All the model errors are within the acceptable range.
Samples of Osmanthus fragrans Lour, Ceder deodara G.Don and Ligustrum lucidum Ait were collected in Xianlin district of Nanjing. The magnetic parameters of tree leaf samples were measured. The correlations between concentrations of atmospheric particulate matter and magnetic parameter of leaves, as well as meteorological data were analyzed. The support vector machine(SVM) method was used to construct the magnetic evaluation model of atmospheric particulate matter. Results showed that the value of mass magnetization rate(χ), saturated isothermal remanence(SIRM) and anhysteretic remanent magnetization(ARM) of leaves of cedar were the highest. The three leaves accumulated relatively more fine-magnetic domain magnetic particles in summer and autumn, whereas relatively more coarse-magnetic domain magnetic particles in spring and winter. PM_(10) and PM_(2.5) concentrations showed significant negative correlation with wind speed, temperature, and humidity, and significant positive correlation with atmospheric pressure. The magnetic parameters of three species of leaves had positively correlations with PM_(10) and PM_(2.5) concentrations(p<0.05). The correlation coefficients(r) of magnetic parameters of tree leaves were higher with χ and SIRM than with ARM.The correlation coefficient(r) of magnetic parameters of Osmanthus fragrans Lour leaves with PM_(10) and PM_(2.5) concentrations were both higher than 0.560. Taking "leaf magnetic parameters and meteorological data" as the input factors, the simulation effect of SVM models on PM_(2.5) concentration was better, while the simulation effect of taking only "leaf magnetism parameter" as the input factor on PM_(10) concentration was better. All the model errors are within the acceptable range.
引文
[1] 平措.我国城市大气污染现状及综合防治对策[J].环境科学与管理,2006,31(1):18-21.
[2] Hanesch M,Scholger R,Rey D.Mapping dust distribution around an industrial site by measuring magnetic parameters of tree leaves [J].Atmospheric Environment,2003,37(36):5125-5133.
[3] Jordanova D,Jordanova N,Petrov P.Magnetic susceptibility of road deposited sediments at a national scale--relation to population size and urban pollution.[J].Environmental Pollution,2014,189(12):239-251.
[4] 姜月华,殷鸿福,王润华.环境磁学理论、方法和研究进展[J].地球学报,2004,25(3):357-362.
[5] Monaci F,Moni F,Lanciotti E,et al.Biomonitoring of airborne metals in urban environments:new tracers of vehicle emission,in place of lead.[J].Environmental Pollution,2000,107(3):321-327.
[6] Alfani A,Baldantoni D,Maisto G,et al.Temporal and spatial variation in C,N,S and trace element contents in the leaves of Quercus ilex within the urban area of Naples.[J].Environmental Pollution,2000,109(1):119.
[7] Hanesch M,Scholger R,Rey D.Mapping dust distribution around an industrial site by measuring magnetic parameters of tree leaves[J].Atmospheric Environment,2003,37(36):5125-5133.
[8] Urbat M,Lehndorff E,Schwark L.Biomonitoring of air quality in the Cologne conurbation using pine needles as a passive sampler—Part I:magnetic properties[J].Atmospheric Environment,2004,38(23):3781-3792.
[9] 陈俏,曹根牛,陈柳.支持向量机应用于大气污染物浓度预测[J].计算机技术与发展,2010,20(1):250-252.
[10] 李勇,白云,李川.大气污染物SO2预测模型研究综述[J].四川环境,2016,35(1):144-148.
[11] Su,Rez S,Nchez A,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 An International Journal,2011,54(5-6):1453-1466.
[12] 张建磊,乐群,束炯.最小二乘支持向量机在臭氧浓度时间序列预测中试应用[J].江苏环境科技,2007,20(3):43-45.
[13] 冷湘梓,王勤耕,钱新,等.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.
[14] 常涛.支持向量机在大气污染预报中的应用研究[J].气象,2006,32(12):61-65.
[15] 王呈,王金花,钱新,等.南京市树叶附尘对大气重金属污染的磁学响应[J].中国环境科学,2017,37(7):2414-2423.
[16] 俞立中,许羽.湖泊沉积物的矿物磁性测量及其环境应用[J].地球物理学进展,1995,10(1):11-22.
[17] 周元涛,张玉芬.环境磁学及其在古气候环境研究中的应用[J].工程地球物理学报,2007,4(6):533-539.
[18] 张春霞,黄宝春.环境磁学在城市环境污染监测中的应用和进展[J].地球物理学进展,2005,20(3):705-711.
[19] 陈满荣,王少平,俞立中.环境磁学及其在地理环境研究中的应用[J].云南地理环境研究,2001,13(1):11-19.
[20] 孙知明,胡守云,马醒华.现代湖泊沉积物中磁性矿物的研究及其环境意义[J].地球物理学报,1996,(2):178-187.
[21] 胡守云,段雪梅,沈明洁,等.北京首钢工业区大气重金属污染树叶的磁学响应[J].科学通报,2008,53(4):437-445.
[22] 刘飞,褚慧敏,郑祥民.上海市公园绿地树叶对大气重金属污染的磁学响应[J].环境科学,2015,36(12):4374-4380.
[23] 隆茜,周菊珍,孟颉,等.城市道路绿化带不同植物叶片附尘对大气污染的磁学响应[J].环境科学,2012,33(12):4188-4193.
[24] Cao L,Appel E,Hu S,et al.Magnetic response to air pollution recorded by soil and dust-loaded leaves in a changing industrial environment [J].Atmospheric Environment,2015,119(12):304-313.
[25] 曹丽婉,胡守云,Appel,等.临汾市树叶磁性的时空变化特征及其对大气重金属污染的指示[J].地球物理学报,2016,59(5):1729-1742.
[26] 李勇.不同环境中树叶的磁学性质及其对大气环境污染的指示[J].现代地质,2016,30(3):606-613.
[27] 彭舜磊,李鹏,李世春,等.煤炭型城市绿化树种叶片磁性及对PM2.5的指示作用[J].环境科学与技术,2017,(9):38-42.
[28] 张春霞,黄宝春,李震宇,等.高速公路附近树叶的磁学性质及其对环境污染的指示意义[J].科学通报,2006,51(12):1459-1468.
[2] Hanesch M,Scholger R,Rey D.Mapping dust distribution around an industrial site by measuring magnetic parameters of tree leaves [J].Atmospheric Environment,2003,37(36):5125-5133.
[3] Jordanova D,Jordanova N,Petrov P.Magnetic susceptibility of road deposited sediments at a national scale--relation to population size and urban pollution.[J].Environmental Pollution,2014,189(12):239-251.
[4] 姜月华,殷鸿福,王润华.环境磁学理论、方法和研究进展[J].地球学报,2004,25(3):357-362.
[5] Monaci F,Moni F,Lanciotti E,et al.Biomonitoring of airborne metals in urban environments:new tracers of vehicle emission,in place of lead.[J].Environmental Pollution,2000,107(3):321-327.
[6] Alfani A,Baldantoni D,Maisto G,et al.Temporal and spatial variation in C,N,S and trace element contents in the leaves of Quercus ilex within the urban area of Naples.[J].Environmental Pollution,2000,109(1):119.
[7] Hanesch M,Scholger R,Rey D.Mapping dust distribution around an industrial site by measuring magnetic parameters of tree leaves[J].Atmospheric Environment,2003,37(36):5125-5133.
[8] Urbat M,Lehndorff E,Schwark L.Biomonitoring of air quality in the Cologne conurbation using pine needles as a passive sampler—Part I:magnetic properties[J].Atmospheric Environment,2004,38(23):3781-3792.
[9] 陈俏,曹根牛,陈柳.支持向量机应用于大气污染物浓度预测[J].计算机技术与发展,2010,20(1):250-252.
[10] 李勇,白云,李川.大气污染物SO2预测模型研究综述[J].四川环境,2016,35(1):144-148.
[11] Su,Rez S,Nchez A,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 An International Journal,2011,54(5-6):1453-1466.
[12] 张建磊,乐群,束炯.最小二乘支持向量机在臭氧浓度时间序列预测中试应用[J].江苏环境科技,2007,20(3):43-45.
[13] 冷湘梓,王勤耕,钱新,等.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.
[14] 常涛.支持向量机在大气污染预报中的应用研究[J].气象,2006,32(12):61-65.
[15] 王呈,王金花,钱新,等.南京市树叶附尘对大气重金属污染的磁学响应[J].中国环境科学,2017,37(7):2414-2423.
[16] 俞立中,许羽.湖泊沉积物的矿物磁性测量及其环境应用[J].地球物理学进展,1995,10(1):11-22.
[17] 周元涛,张玉芬.环境磁学及其在古气候环境研究中的应用[J].工程地球物理学报,2007,4(6):533-539.
[18] 张春霞,黄宝春.环境磁学在城市环境污染监测中的应用和进展[J].地球物理学进展,2005,20(3):705-711.
[19] 陈满荣,王少平,俞立中.环境磁学及其在地理环境研究中的应用[J].云南地理环境研究,2001,13(1):11-19.
[20] 孙知明,胡守云,马醒华.现代湖泊沉积物中磁性矿物的研究及其环境意义[J].地球物理学报,1996,(2):178-187.
[21] 胡守云,段雪梅,沈明洁,等.北京首钢工业区大气重金属污染树叶的磁学响应[J].科学通报,2008,53(4):437-445.
[22] 刘飞,褚慧敏,郑祥民.上海市公园绿地树叶对大气重金属污染的磁学响应[J].环境科学,2015,36(12):4374-4380.
[23] 隆茜,周菊珍,孟颉,等.城市道路绿化带不同植物叶片附尘对大气污染的磁学响应[J].环境科学,2012,33(12):4188-4193.
[24] Cao L,Appel E,Hu S,et al.Magnetic response to air pollution recorded by soil and dust-loaded leaves in a changing industrial environment [J].Atmospheric Environment,2015,119(12):304-313.
[25] 曹丽婉,胡守云,Appel,等.临汾市树叶磁性的时空变化特征及其对大气重金属污染的指示[J].地球物理学报,2016,59(5):1729-1742.
[26] 李勇.不同环境中树叶的磁学性质及其对大气环境污染的指示[J].现代地质,2016,30(3):606-613.
[27] 彭舜磊,李鹏,李世春,等.煤炭型城市绿化树种叶片磁性及对PM2.5的指示作用[J].环境科学与技术,2017,(9):38-42.
[28] 张春霞,黄宝春,李震宇,等.高速公路附近树叶的磁学性质及其对环境污染的指示意义[J].科学通报,2006,51(12):1459-1468.
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |