Chapter three: methodology of exposure modeling

In this chapter, the concept of exposure assessment and its evolution is introduced, and evaluated by critically appraising the pertinent literature as it applies to exposures to Particulate Matter (PM). Exposure measurement or estimation methodologies and models are reviewed. Three exposure/measurement methodologies are assessed. Estimation methods focus on source evaluation and attribution, sources include those outdoors and indoors as well as in occupational and in-transit environments. Fate and transport models and their inputs are addressed to estimate concentrations outdoors and indoors; source attribution techniques help focus on the contributing sources. Activity pattern techniques are also reviewed and their use in exposure models to estimate inhalation exposure to PM is presented. Deterministic, regression and other stochastic models of exposure to PM are reviewed and evaluated. Strengths, limitations, assumptions and affirmations of the use of exposure assessment as an integral component of risk assessment and risk management are discussed in the conclusions and discussions section of this work.     

PM2.5 chemical source profiles for vehicle exhaust, vegetative burning, geological material, and coal burning in Northwestern Colorado during 1995

PM_2.5 (particles with aerodynamic diameters less than 2.5 μm) chemical source profiles applicable to speciated emissions inventories and receptor model source apportionment are reported for geological material, motor vehicle exhaust, residential coal (RCC) and wood combustion (RWC), forest fires, geothermal hot springs; and coal-fired power generation units from northwestern Colorado during 1995. Fuels and combustion conditions are similar to those of other communities of the inland western US. Coal-fired power station profiles differed substantially between different units using similar coals, with the major difference being lack of selenium in emissions from the only unit that was equipped with a dry limestone sulfur dioxide (SO₂) scrubber. SO₂ abundances relative to fine particle mass emissions in power plant emissions were seven to nine times higher than hydrogen sulfide (H₂S) abundances from geothermal springs, and one to two orders of magnitude higher than SO₂ abundances in RCC emissions, implying that the SO₂ abundance is an important marker for primary particle contributions of non-aged coal-fired power station contributions. The sum of organic and elemental carbon ranged from 1% to 10% of fine particle mass in coal-fired power plant emissions, from 5% to 10% in geological material, >50% in forest fire emissions, >60% in RWC emissions, and >95% in RCC and vehicle exhaust emissions. Water-soluble potassium (K⁺) was most abundant in vegetative burning profiles. K⁺/K ratios ranged from 0.1 in geological material profiles to 0.9 in vegetative burning emissions, confirming previous observations that soluble potassium is a good marker for vegetative burning.     

Second generation chemical mass balance source apportionment of sulfur oxides and sulfate at the Grand Canyon during the Project MOHAVE summer intensive

Receptor-based chemical mass balance (CMB) analysis techniques are designed to apportion species that are conserved during pollutant transport using conserved source profiles. The techniques will fail if non-conservative species (or profiles) are not properly accounted for in the CMB model. The straightforward application of the CMB model developed for Project MOHAVE using regional profiles resulted in a significant under-prediction of total sulfate oxides (SOx, SO2 plus fine particulate sulfate) for many samples at Meadview, AZ. In addition, for these samples the concentration of the inert tracer emitted from the MOHAVE Power Project (MPP), ocPDCH, was also under-predicted. A second-generation model has been developed which assumes that separation of particles and SO2 can occur in the MPP plume during nighttime stable plume conditions. This second-generation CMB model accounts for all SOx present at the various receptor sites. In addition, the concentrations of ocPDCH and the presence of other inert tracers of emission from regional sources are accurately predicted. The major source of SOx at Meadview was the MPP, but the major source of sulfate at this site was the Las Vegas urban area. At Hopi Point in the Grand Canyon, the Baja California region (Imperial Valley and northwestern Mexico) was the major source of both SOx and sulfate.     

On-road vehicle particulate matter and gaseous emission distributions in Las Vegas, Nevada, compared with other areas

During the spring and summer of 2000, 2001, and 2002, gaseous and particulate matter (PM) fuel-based emission factors for approximately 150,000 low-tailpipe, individual vehicles in the Las Vegas, NV, area were measured via on-road remote sensing. For the gaseous pollutants (carbon monoxide, hydrocarbons, and nitrogen oxide), a commercial vehicle emissions remote sensing system (VERSS) was used. The PM emissions were determined using a Lidar-based VERSS. Emission distributions and their shapes were analyzed and compared with previous studies. The large skewness of the distributions is evident for both gaseous pollutants and PM and has important implications for emission reduction policies, because the majority of emissions are attributed to a small fraction of vehicles. Results of this Las Vegas study and studies at other geographical locations were compared. The gaseous pollutants were found to be close to those measured by VERSS in other U.S. cities. The PM emission factors for spark ignition and diesel vehicles are in the range of previous tunnel and dynamometer studies.     

环境条件对遵义市呼吸系统疾病的影响及预测研究

为探究气象和空气污染等环境条件对呼吸系统疾病发病的影响,为遵义地区相关疾病预防提供科学依据,采用分布滞后非线性模型和广义线性、相加模型,利用当地气象和污染资料,分析了2012~2016年遵义市气象环境要素对呼吸系统疾病发病的影响.结果显示,遵义市呼吸系统疾病发病状况与当地长期气候状态基本保持一致,气候效应对其的影响占主导作用,其中,冬季为呼吸系统疾病高发期,立秋至处暑前后其发病人数最少,表明此时间段内当地气候条件对呼吸系统疾病患者有气候疗养效应.气温对呼吸系统疾病发病的影响以低温滞后效应为主,在其敏感阈值附近气温每变化1℃,发病人数将累积增加31.6%（95% CI：4.4%~65.8%）;气压以高压滞后效应为主,相对湿度则在低湿部分同时有即时和滞后效应.舒适度对呼吸系统疾病的影响,在冷、热斜胁迫下其发病人数明显多于舒适状况时.PM_2.5、SO₂和NO₂三种污染物的影响都以即时效应为主,而CO则在累积滞后lag04时相对危险度最高,PM_2.5与呼吸系统疾病发病人数的暴露-反应曲线呈单调线性分布,SO₂、NO₂和CO均为“J”型分布.低温与高浓度NO₂或者低湿与高浓度SO₂的协同作用对呼吸系统疾病的影响较大.建立的全年和季节多元逐步回归方程的试预报准确率在75%以上（夏季除外）,其中分季节建模预测效果显著优于全年预测效果.     

研究星载IPDA激光雷达探测CO2光谱纯度

光谱纯度是积分路径差分吸收激光雷达最重要的系统参数之一,光谱纯度直接影响CO₂数据的探测精度.在模式研究中要求输入观测的CO₂数据误差小于1×10^-6,对激光器光谱纯度参数设计提出很高要求.本文采用有效吸收截面分析探测反演误差的方法,研究了由光谱纯度带来的CO₂柱浓度探测误差,并基于窄带滤波器对光谱不纯能量的抑制作用,对不同带宽的窄带滤波器进行选择和分析,从而达到降低对光谱纯度的要求,提升探测精度的目的,最后讨论了由于窄带滤波器造成的能量衰减所对随机误差的影响.结果表明：当光谱纯度为99.9%,窄带滤波器带宽1GHz,透射率为0.86,脉冲能量为100mJ时CO₂探测的系统误差小于0.084×10^-6,随机误差小于0.02×10^-6,满足探测精度要求.     

锰矿石人工湿地中去除双氯芬酸的机理研究

以双氯芬酸为目标有机药物污染物,探究其在以锰矿石为填料并接种了异化金属还原菌（Geobactermetallireducens,简称GS-15细菌）的垂直流人工湿地中的去除过程.结果表明,双氯芬酸在锰矿石人工湿地中的平均去除率最高为23.56%.采用XRD和XRF对反应前后的锰矿石进行物相和元素相对含量分析,得出锰（Mn）是参与降解双氯芬酸的关键响应元素.GS-15细菌利用锰矿石发生异化还原对双氯芬酸的氧化降解是锰矿石人工湿地对双氯芬酸的主要去除途径.此外,利用液质联用技术（LC-MS/MS）对降解产物进行鉴定,发现双氯芬酸被降解生成了5-羟基双氯芬酸、双氯芬酸-2,5-亚氨基琨和1,3-二氯苯三种降解产物,该研究成果对有机药物废水的深度处理提供新的方法和理论指导.     

沈阳市PM2.5离子成分对呼吸疾病门诊数影响研究

采用时间序列的半参数广义相加模型,在控制了长期趋势、"星期几效应"和气象因素等混杂因素的基础上,分析沈阳市大气污染物及PM_2.5中水溶性离子对呼吸系统疾病门诊就诊人数的影响,并按性别和年龄分层建模.结果表明：PM_2.5及其各离子成分与呼吸系统疾病门诊人数之间存在关联,并有明显的滞后效应.受冬季供暖燃煤排放影响,PM_2.5、NO₃^-和NH₄⁺呈显著关联,在滞后累积2d后风险最大.最佳滞后时间下,PM_2.5的浓度每增加10µg/m³,对应呼吸系统疾病日门诊就诊人数增加百分比（ER）为1.31%（95% CI：1.2%~1.43%）;离子成分SO₄^2-、NO₃^-、NH₄⁺、Cl^-、K⁺、Mg²⁺、Ca²⁺和Na⁺的浓度每增加1个4分位间距（IQR）,对应的呼吸系统疾病日门诊就诊人数增加百分比（ER）分别为3.22%（95% CI：2.81%~3.62%）、4.67%（95% CI：4.13%~5.22%）、5.41%（95% CI：4.49%~6.33%）、7.38%（95% CI：3.91%~10.96%）、0.14%（95% CI：-6.34%~7.07%）、7.64%（95% CI：-11.87%~31.47%）、3.57%（95% CI：-2.83%~10.39%）和0.46%（95% CI：-16.64%~21.06%）.PM_2.5、Cl^-、Mg²⁺、Ca²⁺和Na⁺对女性呼吸疾病门诊人数的影响比对男性的影响大.PM_2.5、SO₄^2-、Cl^-、Ca²⁺和Na⁺对≥65岁的老人门诊人数的影响比对15~65岁劳动年龄人群的影响大.表明不同性别、不同年龄由于生理结构和环境因素的不同而引起的差异不同.     

1951~2014年中国北方地区气温突变与变暖停滞的时空变异性

基于1951~2014年中国北方及周边地区357个气象站点平均最低气温、平均气温和平均最高气温年（月）数据,采用M~K检验等方法,分析了中国北方地区3类气温突变和变暖停滞特征的时空变异性.结果表明：研究区3类气温整体突变年（1978~1999年、1981~2002年、1981~2005年）、分布广泛的普遍突变年（1988年、1989年、1997年）及范围（3a）均依次变晚.整体上,突变年随纬度降低变晚,东北突变早于西北和华北地区.变暖停滞集中于1998和2007年及其前后,3类气温亦依次变晚（1994~2007年、1995~2009年、1998~2010年）,由黄河流域中段向其他方向越来越晚.突变至变暖停滞周期整体随纬度降低缩短（3~30a）,突变越早周期越长.西北地区突变与变暖停滞前后各时段均值温差最大（2.4℃）,温差在1℃左右站点分布最广泛.各时段升（降）温速率整体依次在0.01℃/10a、0.05℃/10a、-0.03℃/10a左右站点分布最广泛,突变后升温最快（0.02~0.16℃/10a）,且西北地区对升温贡献最大,变暖停滞后东北地区对降温贡献最大,2时段按平均最低气温、平均最高气温、平均气温顺序升（降）温速率递减.3类气温波动程度减弱,整体随纬度降低.高纬度、高海拔和山地地区突变和变暖停滞较周边地区偏早或偏晚,特征值较大.整个北方地区3类气温突变、变暖停滞、突变与变暖停滞时间及各时段特征值各自具有自身一致性的普遍规律.     

南方林区2000~2016年林火释放污染物动态变化研究

以南方林区为研究区域,基于林业统计年鉴和MODIS卫星火点数据,结合相关文献中不同林型各类污染物的排放因子数据,估算2000~2016年间研究区域各省森林生物质燃烧及污染物排放总量,并分析其时空变化情况.结果表明,南方林区森林火灾总体呈先升后降的趋势,林火空间分布较为分散.2000~2016年南方林区森林生物质燃烧总量4.79×10⁴kt.其中,温带常绿针叶林、温带常绿阔叶林、温带落叶阔叶林、热带常绿阔叶林和灌木的燃烧比例分别为65.24%,17.45%,1.55%,1.10%和14.66%.CO、CO₂、NO_x、CH₄、VOCs和PM_2.5等主要污染物的排放总量均值依次为3.73×10³,5.87×10⁴,48.05,347.27,358.54和365.06kt.此外,各类污染物的时空分布差异性明显：福建、广东、广西全省,海南西部,以及江西和湖南南部是各类污染物网格排放高值区;江西和湖北各类污染物的排放呈显著下降,福建NO_x、VOCs和PM_2.5,以及广东CH₄和VOCs下降趋势显著,其余省份各类污染物排放下降但不显著.林火释放污染物占工业粉尘比例呈波动性上升,表明该地区森林火灾对大气环境的影响程度逐渐增加.