• 关键词：大气PM₁₀  天气类型  背景时段  取值概率

• 基金项目：北京市科委项目(No.H030630540320)

• 李金香
• 1. 北京市环境保护监测中心 北京 100044; 2. 中国科学院安徽光学精密机械研究所 合肥 230031

• 虞统
• 北京市环境保护监测中心 北京 100044

• 赵越
• 北京市环境保护监测中心 北京 100044

• 李海军
• 北京市环境保护监测中心 北京 100044

• 李令军
• 北京市环境保护监测中心 北京 100044

• 刘文清
• 中国科学院安徽光学精密机械研究所 合肥 230031

• 摘要：为了消除或降低PM₁₀环境背景值计算中城市局地污染的影响,并且提高背景值在污染过程模拟中的准确程度,紧密联系天气过程、通过监测和统计途径探讨了确定环境背景值的方法.北京市空气质量背景监测站点在地方性天气过程中明显受到城市污染的影响,文中根据天气形势场、垂直方向上风向一致性、系统影响持续时间等判别标准区分了系统性和地方性过程,筛选出受系统性过程影响的、"背景时段"的监测数据.背景时段监测数据统计计算时,将背景值视为变量,用百分位数、剪裁均值、最大概率密度值、50%概率区间等特征统计量来表达其概率分布.文中还将背景值细化,根据天气系统特征和区域自然环境、人类活动的差异将背景值区分为4类,以代表不同污染过程的背景水平.使用2003年11月至2004年12月北京市北部和南部共5个清洁站点的监测数据,计算了117个中尺度天气系统影响时,PM₁₀浓度概率密度分布以及背景值特征统计量.结果表明,4类背景值差别很大,从小到大排序依次是:北风 >东风 >西风 >南风,剪裁均值依次为:24～34、58、89～110和105μg·m^-3 .对4类背景值综合运算,得到2004年PM₁₀年尺度上的综合背景值为58～67μg·m^-3 ,较好地扣除了局地污染的影响.

• Abstract：Asynoptic-type-based approach is proposed and examined in order to reduce the contribution of local emissions to urban atmospheric environmental background and to improve the accuracy of estimating PM₁₀ background values when modeling a single pollution process.Background measurements are greatly influenced by local emissions in topography-induced weather processes.Therefore,Background Periods',which are identified by meso-scale synoptic processes with more than 13h of influencing time,were introduced to select valid measurements and remove local pollution effects.Valid measurements were classified into four types based on the four wind directions to reflect the differences of synoptic processes and regional anthropogenic emissions.As for background concentration,the probability distribution of each type was characterized using percentiles,trimmed mean,and maximum and 50% probability concentration ranges.Finally,PM₁₀ data continuously collected at five suburban sites located in the north and south of Beijing from November,2003 to December,2004 were used to demonstrate the proposed approach.From the data set,117 meso-scale synoptic processes were identified as background periods,and the corresponding observed data were used to calculate PM₁₀ background concentrations.The results show that the trimmed means of PM₁₀ background concentration vary among the 4 groups: 24～34、58、89～110 and 10⁵μg·m^-3 for northern,eastern,western and southern wind groups,respectively.By averaging the measurements of the 4 groups,the annual PM₁₀ background concentration for the year of 2004 was calculated to be 58～67μg·m^-3 .Comparison with other approaches indicates that this approach can remove much local pollution and consequently obtain more accurate background values of PM₁₀.

• Key words：atmospheric PM₁₀  synoptic type  background periods  probability

• 
  摘要点击次数： 2026 全文下载次数： 2985