臺灣大學: 環境工程學研究所張能復; 余泰毅盧永勝Lu, Yong-ShengYong-ShengLu2013-04-102018-06-282013-04-102018-06-282010http://ntur.lib.ntu.edu.tw//handle/246246/257000近年來在雲嘉南地區的臭氧污染有逐年惡化的趨勢，因此本研究 針對1999?2009年期間，以多變量統計方法及光化學軌跡模式分析近 年來空氣品質監測資料，深入瞭解各重大污染源對於雲嘉南地區臭氧 污染的貢獻，提供臭氧管制策略之參考依據。 本研究首先探討雲嘉南地區汙染物的時空變異情形，結果顯示 1999?2009 年雲嘉南地區NOx 與NMHC 有下降的趨勢，而台南地 區因為季節風向的關係，使得秋季的臭氧站日數較多。進而利用成分 分數大於2.5 的標準，篩選出有代表性的臭氧事件日，因為秋季雲嘉 南地區暴露在長時間低臭氧濃度下較嚴重，與非事件日的比較中得知 若是處於氣象條件不利擴散的情況，就容易出現臭氧的尖峰值，在軌 跡線通過汙染源的情況下，除非擴散條件良好否則就會造成高臭氧的 生成，因此必須管制臭氧前驅物才能有效的削減臭氧的增量。 以相關性分析、因素分析、集群分析三種方法分析雲嘉南空品區 臭氧濃度的相關性與地域分類，並探討各分類區域之特徵。In these years, the pollution of ozone becomes worse in Yunlin, Chiayi, Tainan (Y-C-T) area in Taiwan. This study used multivariate statistical analysis and trajectory photochemical air quality model to analyze the database of air quality from 1999 to 2009. The goal of this study is to understand the pollution magnitude produced by sources of pollution, and provided the basis for control of ozone. This study investigated that the change on spatial and temporal variations of (Y-C-T) area. The result indicated that the NOx and NMHC had decreased in 1999 to 2009, but the ozone pollution was still worse in fall.Tainan station due to the direction of the monsoon that cause more of ozone hour station days. Factor scores above 2.5 is a useful index for selection of ozone episodes. Compare with non-ozone episodes, the result indicated that the highest concentration of ozone was produced by source of pollution when the climate was not conductive to diffuse. For decrease the magnitude of ozone, ozone precursors should be controlled. This study applied three statistical methods including correlation analysis, factor analysis, cluster analysis to investigate similarities of pollutant levels for monitoring stations and area classification in (Y-C-T) area,and investigated characteristics of those seprational regions.14645624 bytesapplication/pdfen-US臭氧事件日多變量統計光化學軌跡模式相關性分析因素分析集 群分析主成分分析成分分數ozoneepisodemultivariate statistical analysistrajectory photochemical air quality modelcorrelation analysisfactor analysiscluster analysisprincipal component analysisfactor score[SDGs]SDG13thesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/257000/1/ntu-99-R97541131-1.pdf