摘要
为了研究大气中PM2.5污染特征以及其随时间变化规律,基于西安市2013年1月—2014年4月间SO2、NO2、CO、O3、日最高温度(Tmax)、日最低温度(Tmin)、PM2.5、PM10等因素的监测数据。运用统计学原理和多元回归分析方法,分析了PM2.5的污染特征及相关因素对其产生的贡献度,进一步建立了四季的最优多元回归模型。研究结果表明,西安市年平均质量浓度124.9μg/m3,四季的平均污染浓度从大到小依次为冬、春、秋、夏;春夏两季贡献较大的为SO2、CO;秋冬两季贡献较大的为NO2、CO;最终建立的模型的相关系数较高,模型很好地拟合了冬春两季PM2.5变化趋势,能较准确地反映了西安市PM2.5的污染特征,具有一定的理论和实用价值。
The pollution characteristics of PM2. 5and its time-dependent variation were studied based on the monitoring data of the daily average concentration of SO2,NO2,CO,O3,PM2. 5,PM10 and the city's daily maximum and minimum temperature in Xi'an from January 2013 to April 2014. Statistical theory and multivariate regression analysis were firstly employed to analyze the pollution characteristics of PM2. 5and the contribution of related factors. An optimal multivariate regression model for four seasons was then established. The result showed that the annual average mass concentration of PM2. 5in Xi'an is 124. 9 μg / m3,and the average pollution concentration is in a descending order of winter,spring,autumn and summer. Besides,the result indicated that SO2 and CO serve as the principal pollutants in spring and summer,while NO2 and CO contribute more in autumn and winter. The regression model possesses a comparatively high correlation coefficient and has a certain theoretical and practical value. It well fits the variation trend of PM2. 5in winter and accurately reflects the pollution characteristics of PM2. 5in Xi'an.
出处
《环境工程学报》
CAS
CSCD
北大核心
2015年第6期2974-2978,共5页
Chinese Journal of Environmental Engineering
关键词
PM2.5
污染特征
多元回归模型
变化趋势
拟合
PM2.5
pollution characteristics
multivariate regression model
variation trend
fitting
