Hybrid Neural Networks and Boosted Regression Tree Models for Predicting Roadside Particulate Matter

This paper examines the application of artificial neural network (ANN) and boosted regression tree (BRT) methods in air quality modelling. The methods were applied to developing air quality models for predicting roadside particle mass concentration (PM₁₀, PM_2.5) and particle number counts (PNC) based on air pollution, traffic and meteorological data from Marylebone Road in London. Elastic net, Lasso and principal components analysis were used as feature selection methods for the ANN models to reduce the number of predictor variables and improve their generalisation. The performance of the ANN with feature selection (ANN hybrid) and the BRT models was evaluated and compared using statistical performance metrics. The performance parameters include root mean square error (RMSE), fraction of prediction within a factor of two of the observation (FAC2), mean bias (MB), mean gross error (MGE), the coefficient of correlation (R) and coefficient of efficiency (CoE) values. The input variables selected by the elastic net produced the best performing ANN models. The ANN hybrid produced models performed only slightly better than the BRT models. The R values of the ANN elastic net and BRT models were 0.96 and 0.95 for PM₁₀, 0.96 and 0.96 for PM_2.5 and 0.89 and 0.87 for PNC, respectively. Their corresponding CoE values were 0.72 and 0.70 for PM₁₀, 0.74 and 0.76 for PM_2.5 and 0.81 and 0.71 for PNC respectively. About 80–99% of all the model predictions are within a factor of two of the observed particle concentrations. The BRT models offer more advantages regarding model interpretation and permit feature selection. Therefore, the study recommends the use of BRT over ANN where the model interpretation is a priority.     

2001年～2008年及奥运会期间天津市大气污染特征分析

根据天津市大气质量监测数据,对2001年～2008年及奥运会期间天津市大气污染特征和主要大气污染物的变化规律进行了分析。结果表明,2001年～2008年天津市的PM10、SO2和NO2污染总体呈下降趋势,但质量浓度仍相对较高。2008年8月奥运会期间天津市PM10和SO2质量浓度达到国家空气质量二级标准,NO2质量浓度达到国家空气质量一级标准,空气质量良好。天津市PM10污染相对稳定,SO2和NO2的污染分布呈现明显的季节性,时间上表现为冬强夏弱。气象条件对污染物浓度影响明显,沙尘、大雾等天气可使污染物浓度急剧升高。     

烟花爆竹禁燃对春节期间东北地区空气质量影响评估

利用2015—2017年春节期间东北地区主要大气污染物(PM_(10)、PM_(2.5)、SO_2、NO_2、CO和O3)质量浓度监测资料及相应气象因子(温度、湿度、风速和气压)观测资料,分析了春节期间烟花爆竹禁燃对东北地区空气质量的影响。结果表明:随着东北地区主要城市禁燃力度的增强,空气质量逐年提升,PM_(2.5)和SO_2浓度逐年大幅度下降。禁燃可明显降低城区PM_(2.5)浓度,而由于春节期间污染源整体减少,城区和城郊监测点PM_(2.5)浓度值差异减小。烟花爆竹对PM_(10)和PM_(2.5)浓度影响高于对气体污染物SO_2、NO_2和CO的影响。此外,气象条件对东北地区春节期间禁燃改善空气质量的效果也有明显影响。因此,结合春节期间的气象条件,在东北地区实施禁燃政策动态调整非常必要。     

黑龙江省环境空气污染物和气象要素的典型相关分析

通过对黑龙江省4个自然年(2016年1月1日—2019年12月31日)环境空气污染物和气象要素的分析,揭示了黑龙江省气象条件对空气污染物浓度的影响规律与特征。对PM_(2.5)、PM_(10)、SO_2、NO_2、CO和O_3等6项污染物的描述性统计和简单的相关分析显示:黑龙江省环境空气质量呈现逐年变好的趋势,非采暖期环境空气质量好于采暖期,6项污染物中除O_3呈现夏季偏高以外,其余污染物采暖期浓度均高于非采暖期。运用典型相关分析法探究环境空气污染物与温度、降水量、相对湿度、风速和气压5项气象要素之间的关系,并进行统计学检验,结果表明:环境空气污染物与气象要素之间存在显著相关,温度、风速和相对湿度对污染物具有显著影响。非采暖期大气相对湿度对PM_(10)和O_3-8h的影响显著;而在采暖期,风速对PM_(10)和PM_(2.5)的影响显著。     

北京地区不同季节PM2.5和PM10浓度对地面气象因素的响应

利用2013年1月—2014年12月北京地区PM_(2.5)和PM_(10)监测数据和同期近地面气象观测数据,采用非参数分析法(Spearman秩相关系数)研究了北京地区PM_(2.5)和PM_(10)的浓度对不同季节地面气象因素的响应。结果表明:北京地区大气颗粒物浓度水平具有明显的季节特征,冬季大气颗粒物污染最严重,夏季最轻。不同季节影响颗粒物浓度水平的气象因素各不相同,其中风速和日照时数为主要影响因素。PM_(2.5)和PM_(10)质量浓度对气象因素变化的响应程度也有较大区别,PM_(2.5)/PM_(10)比值冬季最高,PM_(2.5)影响最大,春季最低,PM_(10)影响最大。这些结论可对制订科学有效的大气污染控制策略提供参考。     

2021年青岛市一次PM_(2.5)和沙尘混合空气污染过程分析

以2021年3月青岛市空气自动站监测数据为依据,借助环境气象激光雷达、气溶胶激光雷达、在线离子色谱仪等技术手段,并利用后向轨迹模式(HYSPLIT)对青岛市一次PM_(2.5)和沙尘混合空气污染过程、气象条件、颗粒物组成以及传输路径等进行了综合分析。结果表明:静小风、湿度大、垂直方向逆温以及高空多次向近地面的污染物输送是第1阶段PM_(2.5)污染的主要原因,NO^(-)_(3)、SO^(2-)_(4)、NH^(+)_(4)浓度分别占水溶性离子浓度总和的51.7%,24.8%,22.4%,三者之和占ρ(PM_(2.5))的52.3%,机动车源、工业源和燃烧源贡献较大,其中尤以机动车源影响最显著;第2阶段各子站颗粒物浓度变化呈现明显的传输特征,PM_(2.5)中Ca^(2+)浓度升至第1阶段的6倍,沙尘源影响显著,污染气团主要来自蒙古国和我国内蒙古,前期由西北地区直接到达青岛,后期是经渤海湾、烟台到达青岛东南海域,最后回流至青岛;冷高压强度较弱导致近地面水平扩散条件不利,ρ(PM_(10))长时间维持在较高水平。     

Evaluation of the Trend of Air Quality in Taipei, Taiwan from 1994 to 2003

The data collected from the five air quality monitoring stations established by the Taiwan Environmental Protection Administration (TEPA) in Taipei City were analyzed to assess the changes in air quality. The analyses reveal that the air quality in Taipei City improved over the last decade from 1994 to 2003, as evidenced by the significant downward trends of the various primary air pollutant concentrations, such as CO, NO_X, SO₂, and PM₁₀. An air pollution fee was collected by TEPA in 1995, and several air pollution control measures were likewise taken to improve the air quality in Taiwan. However, although the extreme daily maximum O₃ concentrations occurred more frequently in earlier years and showed a downward trend, its moderately high concentrations increased annually in recent years. It implied that after the reduction of various primary pollutant concentrations, the effective reduction of O₃ pollution still remains an important issue.     

沿海城市PM10和PM2.5自动监测过程中质量浓度“倒挂”现象分析与成因探讨

2020年2—3月,位于福建沿海地区中部的莆田市在环境空气质量自动监测过程中出现了严重的PM_(10)和PM_(2.5)质量浓度"倒挂"现象,小时值"倒挂"率为19.86%,日均值"倒挂"率为16.67%。在高相对湿度和低风速气象条件下,颗粒物会出现严重的"倒挂"现象,"倒挂"过程中常伴随着颗粒物和气态污染物(SO_2、NO_2和CO)质量浓度的增加。因此,于2020年2月16日—3月26日开展了颗粒物自动监测和手工监测比对,并结合气象参数、气态污染物质量浓度,以及PM_(10)和PM_(2.5)中水溶性离子和液态水的含量特征,进一步探讨了莆田市颗粒物质量浓度"倒挂"的主要成因。研究表明,PM_(10)和PM_(2.5)自动监测仪器检测原理的差异是导致颗粒物质量浓度"倒挂"的重要原因之一,而气象条件(相对湿度、气温和风速等)、颗粒物质量浓度、颗粒物中主要吸湿组分(NO_3~-、SO_4~(2-)和NH_4~+)和液态水的含量也是颗粒物质量浓度"倒挂"的主要影响因素。莆田市2020年2—3月出现高频率"倒挂"现象是多重因素共同作用的结果,解决该问题需要同时考虑监测仪器检测原理、气象参数、颗粒物质量浓度和吸湿组分等的影响。     

Characterisation of PM10, PM2.5 and Benzene Soluble Organic Fraction of Particulate Matter in an Urban Area of Kolkata, India

In this study, the relationship between inhalable particulate (PM₁₀), fine particulate (PM_2.5), coarse particles (PM_{2.5 – 10}) and meteorological parameters such as temperature, relative humidity, solar radiation, wind speed were statistically analyzed and modelled for urban area of Kolkata during winter months of 2003–2004. Ambient air quality was monitored with a sampling frequency of twenty-four hours at three monitoring sites located near traffic intersections and in an industrial area. The monitoring sites were located 3–5 m above ground near highly trafficked and congested areas. The 24 h average PM₁₀ and PM_2.5 samples were collected using Thermo-Andersen high volume samplers and exposed filter papers were extracted and analysed for benzene soluble organic fraction. The ratios between PM_2.5 and PM₁₀ were found to be in the range of 0.6 to 0.92 and the highest ratio was found in the most polluted urban site. Statistical analysis has shown a strong positive correlation between PM₁₀ and PM_2.5 and inverse correlation was observed between particulate matter (PM₁₀ and PM_2.5) and wind speed. Statistical analysis of air quality data shows that PM₁₀ and PM_2.5 are showing poor correlation with temperature, relative humidity and solar radiation. Regression equations for PM₁₀ and PM_2.5 and meteorological parameters were developed. The organic fraction of particulate matter soluble in benzene is an indication of poly aromatic hydrocarbon (PAH) concentration present in particulate matter. The relationship between the benzene soluble organic fraction (BSOF) of inhalable particulate (PM₁₀) and fine particulate (PM_2.5) were analysed for urban area of Kolkata. Significant positive correlation was observed between benzene soluble organic fraction of PM₁₀ (BSM10) and benzene soluble organic fraction of PM_2.5 (BSM2.5). Regression equations for BSM10 and BSM2.5 were developed.     

2017年上半年北京市及周边地区空气质量特征分析

依托北京市、廊坊市和保定市高密度的地面空气质量监测、气象要素监测以及PM_2.5化学组分监测和后向轨迹分析等手段，对2017年上半年三地的空气质量进行分析。研究发现：三地中北京市空气质量较好，保定市较差。分污染物来看，保定市SO₂浓度水平明显高于廊坊市和北京市，颗粒物PM₁₀和PM_2.5也呈现保定市最高、北京市最低的规律。从污染物日变化来看，CO、SO₂、NO₂、PM₁₀和PM_2.5呈双峰型分布，O₃呈单峰型分布。从区域整体分布规律来看，PM_2.5和SO₂呈现明显的"南高北低"特征。PM_2.5化学组分分析结果表明：1—4月燃煤对该区域空气质量的影响较大，5—6月机动车排放的影响更为凸显。后向轨迹分析结果表明：在2017年上半年到达北京市的气流中有24%来自于北京市南部，且这些气流多为低空传输，表明区域传输对于北京市空气质量具有一定的影响。