| 摘要: | 懸浮微粒氣動直徑小於或等於2.5 μm (PM2.5)容易進入人體呼吸系統，引起發炎或氣喘造成呼吸系統疾病，對於人體健康影響甚大。目前，世界上各主要國家和組織已陸續制訂PM2.5空氣品質標準以保護公眾健康，我國順應世界潮流也即將公告PM2.5空氣品質標準草案。本文彙整各主要國家和組織訂定標準的程序以及評估標準符合度的方法，檢討台灣現行懸浮微粒空氣品質標準及相關法規，利用2006-2010年環保署PM2.5空氣品質監測數據解析台灣目前空氣品質，並提出台灣未來PM2.5空氣品質標準及評估標準符合度方法。 研究結果顯示，台灣PM2.5濃度呈現由北向南逐漸增加的特徵，春、夏、秋、冬四季各空品區PM2.5連續監測月平均濃度值範圍介於19-50、13-27、15-50、16-64 µg m-3，各空品區逐月累積降雨量變化與PM2.5濃度變化呈現相反趨勢，顯示降雨對PM2.5濃度有重大影響。受東北季風影響期間，大陸冷高壓事件日與沙塵事件日PM2.5平均濃度在北台灣萬里測站為44 µg m-3，平常日為23 µg m-3，從差值推演，東北季風影響期間境外傳輸的污染濃度值約為21 µg m-3，事件日PM2.5 24小時濃度為平常日的1.9倍。仿照美國空品標準評估台灣2008-2010年空品監測站PM2.5 24小時監測濃度的標準符合度時，發現即使是選擇較寬鬆的監測數據累積頻率分布第90%濃度來對應，仍有高達86.4%直轄市、縣（市）和91.7%測站不符合美國空品標準。不過，由於評估美國空品標準採用的標準檢測方法(Federal Reference Method, FRM)獲得的PM2.5檢測濃度低於空品站監測濃度，將空品站監測濃度轉換成等似FRM檢測濃度後，再選擇監測數據累積頻率分布第90%濃度對應評估時，會有大約半數的監測站(51.7%)與縣市(50%)符合美國空品標準。顯示訂定PM2.5空品標準，需同時訂定評估法規符合度的檢測方法和監測數據累積頻率分布百分比對應濃度。 綜合各項因素，本文建議台灣環保署未來參照美國訂定的PM2.5 24小時標準值35 µg m-3與年平均值15 µg m-3作為台灣初期的空氣品質標準值；監測方法以美國FRM檢測方法進行PM2.5濃度檢測，採樣頻率為每三天進行一次監測。對於PM2.5 24小時標準值符合度的評估，建議初期先採用監測數據累積頻率分布第90%濃度來對應，五年後再檢討提升符合度評估為監測數據累積頻率分布第98%濃度的可能性，這種方式可宣示空氣品質標準檢討機制的存在和提供各直轄市、縣（市）當局努力改善PM2.5空氣品質的空間。 Suspended particles with aerodynamic diameter less than or equal to 2.5 μm (PM2.5) are inhalable to human respiratory system to cause lung inflammation and/or asthma and thus cast a great impact on human health. At the present stage, several world’s leading countries and organizations already promulgated their PM2.5 air quality (AQ) standards for the protection of public health. A draft of PM2.5 AQ standard will also be promulgated in Taiwan recently. This study collected the procedure for setting up PM2.5 standards and compliance evaluation methods from leading countries and organizations, reviewed current Taiwan particulate AQ standards and related regulations, analyzed Taiwan’s AQ from PM2.5 AQ monitoring data during 2006-2010, and proposed future Taiwan’s PM2.5 AQ standard and compliance evaluation method. The results show that Taiwan’s PM2.5 concentrations are characterized with an increasing trend from north to south. The monthly PM2.5 averages for spring, summer, fall, and winter in each AQ district ranged from 19-50, 13-27, 15-50, and 16-64 µg m-3, respectively. PM2.5 concentrations are varied inversely with monthly accumulated rainfalls in each AQ district. It indicates a significant effect from rainfalls on PM2.5 concentrations. Under the prevailing northeast monsoon, the PM2.5 average affected by transboundary pollution transport was around 21 µg m-3 deduced from the difference between 44 µg m-3 during continental High and dust transport events and 23 µg m-3 during normal days at the Wan-Li site. This shows PM2.5 24-h average for event days is 1.9-fold over normal days. Following the U.S. AQ standard to evaluate standard compliance for Taiwan’s PM2.5 24-h concentrations during 2008-2010, 86.4% of cities or counties and 91.7% of monitoring sites failed to comply with the U.S. standard even adopting a less stringent 90th percentile of cumulative frequency distribution. However, considering the measurements from the U.S. Federal Reference Method (FRM) coupling with the compliance evaluation of the U.S. AQ standard were substantially lower than the continuous monitoring levels, conversions from continuous monitoring level to FRM equivalence were conducted. There are about half of the monitoring sites (51.7%) and cities or counties (50%) complied with the U.S. AQ standard when the less stringent 90th percentile of cumulative frequency distribution is chosen for the evaluation again. It implies that the measurement method for compliance evaluation and the percentile of cumulative frequency distribution from monitoring data need to be set up simultaneously in addition to the establishment of a PM2.5 AQ standard. In summary, this study suggests Taiwan EPA setting up 35 µg m-3 for PM2.5 24-h standard and 15 µg m-3 for annual standard in reference to the U.S. AQ standard initially. And the measurement method for compliance evaluation is through the usage of the U.S. FRM with an operation on a one-in-three day schedule. For PM2.5 24-h standard compliance evaluation, the 90th percentile of cumulative frequency distribution is recommended to choose initially and possibly raised to the 98th percentile five years later after a review process. This policy not only declares a review mechanism for AQ standard but also provides the authorities of cities and counties allowance for improving their PM2.5 AQ. |
