Acta Scientiarum Naturalium Universitatis Pekinensis ›› 2018, Vol. 54 ›› Issue (5): 1103-1113.DOI: 10.13209/j.0479-8023.2018.031

Previous Articles     Next Articles

Risk Assessment of Exposure to PM2.5 in Beijing Using Multi-Source Data

ZHANG Xiya, HU Haibo   

  1. Institute of Urban Meteorology, China Meteorological Administration, Beijing 100089
  • Received:2017-09-22 Revised:2017-11-29 Online:2018-09-20 Published:2018-09-20
  • Contact: ZHANG Xiya, E-mail: xyazhang(at)


张西雅, 扈海波   

  1. 中国气象局北京城市气象研究所, 北京 100089
  • 通讯作者: 张西雅, E-mail: xyazhang(at)
  • 基金资助:


Through GIS spatial analysis, this study firstly conducts spatial distribution of PM2.5 concentrations using PM2.5 data from 35 automatic air quality monitoring stations in Beijing during the period of 2014–2016. Then population spatial processing is carried out based on DMSP/OLS nighttime light data. On this basis, the authors assess the exposure risk to PM2.5 pollution in Beijing from four aspects: PM2.5 concentration, the characteristic of population exposure, the population exposure intensity, and the population weighted concentration. The results show that 1) high PM2.5 concentrations were mainly distributed in the south, while low concentrations were distributed in the north. There was a good spatial coincidence between the distribution of population exposure to PM2.5 and population distribution, i.e. the densely populated area had high risk of population exposure to PM2.5. 2) During 2014–2016, 100% of population exposed to high PM2.5 yearly mean concentrations (>35 μg/m3) which exceeded the secondary level of Ambient Air Quality Standards (GB 3095–2012), and the ratio of population exposed to 24 hourly mean concentrations (>75 μg/m3) declined over a 3-year period. The share of population exposure to exceeding standard PM2.5 concentration was much higher than those in global average level. 3) The population weighted PM2.5 yearly average concentrations and PM2.5 yearly average concentrations had difference, which is related with exposed population and the distribution of PM2.5 pollution. 4) The distributions of PM2.5 concentration and population are different, so the real impact level on health of human of PM2.5 pollution is different from PM2.5 concentration. Thus, taking the factor of population into account, the risk assessment of exposure to PM2.5 pollution is more accurate.

Key words: PM2.5, population exposure, population weighted, risk assessment


基于2014—2016年的北京地区PM2.5监测数据, 用空间插值法获得北京地区的PM2.5空间分布, 并基于DMSP/OLS夜间灯光数据, 模拟得到北京地区的人口密度空间分布。在此基础上, 从PM2.5浓度空间分布、PM2.5污染的人口暴露特征、PM2.5污染人口暴露强度以及人口加权PM2.5浓度4个方面评估北京地区PM2.5污染暴露风险。结果显示: 1) PM2.5浓度呈现南高北低的空间分布特征, 人口暴露风险空间分布与人口密度空间分布呈现高度的一致性, 即人口密度高的区域, PM2.5污染人口暴露风险也相对较高; 2) 2014, 2015, 2016年北京地区GB3095—2012二级年均浓度标准35 μg/m3的超标人口比例均为100%, 24小时平均浓度标准75 μg/m3的超标人口比例呈逐年显著下降趋势; 3) 2014—2016年北京市人口加权PM2.5年均浓度值与PM2.5年均值均存在差异, 差异度与城市暴露人口和污染情况密切相关; 4) 由于PM2.5污染物浓度空间分布特征与人口密度空间分布特征不同, 北京市PM2.5污染对总体人群的实际影响和健康危害与其平均浓度水平并不相同, 因此考虑人口密度空间分布特征的暴露风险评估比只考虑PM2.5污染物浓度的暴露风险评估更准确。

关键词: PM2.5, 人口暴露, 人口加权, 风险评估

CLC Number: