太原市大气细颗粒物变化趋势与预测模型分析  

Variation trend and forecast model of atmospheric PM2.5 in Taiyuan

作　　者：高宇钊 房瑞玲 李少琼[1] 申宁宁[1] 邓亚敏[1] 马彩云[1] 刘桂芬[1] 

机构地区：[1]山西医科大学公共卫生学院卫生统计教研室,太原030001

出　　处：《中国药物与临床》2015年第5期603-606,共4页Chinese Remedies & Clinics

基　　金：国家自然科学基金(81172774);太原市环境监测数据与慢性病关系研究(201482)

摘　　要：目的掌握和预警太原市大气中细颗粒物(PM2.5)日浓度的变化趋势,构建更为准确的预警模型,为改进环境质量,促进居民健康提供监测资料评价的方法学依据。方法收集2013年6月至2014年11月太原市逐日PM2.5网报数据,分析太原市18个月以来日报PM2.5的逐月逐日变动趋势,进一步揭示太原市大气中PM2.5日浓度现状;构建时间序列广义回归条件异方差模型[GARCH(1,1)],对太原市未来大气PM2.5浓度及等级进行短期预警。结果太原市大气中PM2.5日浓度变化易受季节的影响,夏秋季浓度低,冬春季浓度高;构建的GARCH(1,1)时序模型,可较好地预测未来短期内太原市大气PM2.5浓度;2014年实测数据与模型预测值分析趋势一致,2014年大气PM2.5浓度预测结果与去年同期相比基本一致。结论太原市大气PM2.5浓度季节波动明显,10月以后逐渐升高,以冬季为甚;太原市大气PM2.5浓度GARCH(1,1)时间序列模型,不仅有助于了解该地大气PM2.5浓度变化规律,且可作为环境大气质量预警的重要指标,预测效果良好,是环境质量监测数据评价的重要方法之一。Objective To learn and forecast the variation trend of the atmospheric （particulate matter 2.5, PM2.5） in Taiyuan, and to establish a more accurate forecasting model, so as to provide methodological evidence for the evaluation of monitoring data that may help improve environmental quality and promote public health. Methods The daily Internet data of PM2.5 in Taiyuan between June 2013 and November 2014 were collected. Then, the day-to-day and month-to-month variation trends of PM2.5 in Taiyuan over the recent 18 months were analyzed, and the cur-rent situation of daily atmospheric PM2.5 concentration in Taiyuan was further investigated. The time-series model us-ing general autoregressive conditional heteroskedasticity [GARCH （1,1）] was established for short-term forecasting of the concentration and level of atmospheric PM2.5 in Taiyuan in the future. Results The concentration change of dai-ly atmospheric PM2.5 in Taiyuan was readily influenced by the seasons, appearing low in summer and autumn, and high in winter and spring. The established GARCH （1,1） time-series model could well forecast the short-term concen-tration of atmospheric PM2.5 in the future. The trend of measured data was consistent with the value of model predic-tion in 2014. The predicted results of atmospheric PM2.5 concentration in 2014 were basically consistent with those in the same period last year. Conclusion The atmospheric PM2.5 concentration in Taiyuan fluctuated dramatically with the seasons, which gradually increases after October, especially in winter. The time-series GARCH （1,1） model of at-mospheric PM2.5 concentration in Taiyuan not only contributes to understanding the changes of atmospheric PM2.5 concentration in Taiyuan, but also can be an important indicator of forecasting environmental air quality. The model analysis can be used as one of the important methods for the data evaluation of environmental quality monitoring.

关 键 词：空气污染物 预测 时间序列 

分 类 号：X513[环境科学与工程—环境工程]