臺灣地區懸浮微粒、臭氧與氣象因子年循環關聯之分析

本研究將1994-2020年間臺灣空品測站日資料進行長時間分析，探討測站之PSI（Pollution Standards Index, 空氣污染指標）或AQI（Air Quality Index, 空氣品質指標）指標值及污染物副指標值的季節變化，作為確認空污好發季節、主要污染物類型以及不同季節之主導污染物差異的分析結果。污染物濃度的年循環除了作為污染物間的季節相關比對外，亦與氣象日資料的長時間平均結果進行比對，以看出季節變化是否具有關聯，並以重分析資料分析全臺及區域氣象因子與污染物間的季節關聯。綜合指標值及副指標值的超標測站比年循環與月平均風場的分析結果顯示，台灣空氣污染集中於秋季中旬至隔年春季中旬，11-4月間以PM10、PM2.5污染為主，整體為東北風風場，10-11月及4-5月則以O3污染為主，整體為弱東風風場。進一步比對台灣空氣污染物間的濃度年循環，PM10、PM2.5與其前驅物之一的NOx具有季節上的相似性，O3與其前驅物NOx則未能看出有相似的季節變化，CO、NOx整體濃度皆直接反映出排放上的季節類似，季節變化則明顯受到風場影響而形成區域差異。至於空氣污染物與氣象因子的季節變化關聯上，降水主要影響固態粒狀空氣污染物如PM10、PM2.5等的濃度季節變化；風場則對於大多數污染物的濃度季節變化有顯著影響，無論是夏季偏南風或是冬季東北風的情況；而日照與O3的濃度季節變化較有關聯，尤其是日照時數與O3的季節變化，然而夏季的偏南風風場則使得O3濃度出現異常低值，因而造成O3的年循環有雙峰的狀況。本研究透過以上的分析，可看出長時間尺度下空氣污染於不同季節的程度差異，以及背後的原因與機制。關鍵字: 臭氧、PM2.5、PM10、PSI、AQI、空氣污染

關鍵字

臭氧； PM2.5 ； PM10 ； PSI ； AQI ；空氣污染

並列摘要

This study applies air quality daily data from air quality stations in Taiwan during the period between 1994 and 2020 for long-term analysis. Pollution Standards Index (PSI) and Air Quality Index (AQI) together with their associated pollutant sub-index are adopted to examine the peak period of air pollution, main pollutants, and dominant pollutants in different seasons. Through analyzing the long-term mean of station-based meteorological observation data and reanalysis gridded data, our research conducts investigations into the annual cycle of various pollutants for examining potential linkages between individual pollutants and clarifying seasonal relationships with meteorological factors in domestic and East Asian regions. Having compared the rate of stations exceeding the PSI, AQI and relevant sub-index threshold during a year with monthly mean observational wind field, this study finds that the air quality in Taiwan deteriorates mainly from the middle of fall through the middle of next spring. From November to April, PM10 and PM2.5 are the major pollutants and the northeasterly wind dominates the same period. Different from the one peak annual cycle pattern found in PM10 and PM2.5, O3 shows two peaks within a year from October to November and from April to May when the weak easterly wind blows over the studied regions. Further examining the long-term mean annual cycle pattern of different air pollutants, this study suggests that both PM10 and PM2.5 have a seasonal relationship with NOx which is a precursor of both PM2.5 and O3; however, there is no clear linkage found between O3 and NOx in terms of seasonal variation pattern. The concentration of CO and NOx reveals strong causalities with their emission sources and can be considerably affected by seasonal wind field, which in turn leads to regional differences in concentration. As for the relationships between individual air pollutants and meteorological factors, seasonal wind field exerts great influences on the concentration of most pollutants regardless of the dominating southerly wind in summer or northeasterly wind in winter. For solid particulate air pollutants such as PM10 and PM2.5, the amount of precipitation in each season mainly determines their concentrations. Compared with the other analyzed meteorological factors, the formation of O3 is more positively related to sunshine especially the sunshine duration which shows a strong seasonal resemblance with O3 concentration. On the other hand, the dominating southerly wind leads to an abnormally low O3 concentration in summer, forming the mentioned double peak patterns in the annual cycle of O3 concentration. Through the discussions above, this study demonstrates the severity of air pollution in different seasons in Taiwan and explains the associated underlying meteorological mechanisms from a long-term perspective. Keywords: Ozone, PM2.5, PM10, PSI, AQI, Air Pollution