台灣中部都市地區有機氣膠之生成及演變過程：以西屯為例

Abstract

有機氣膠（OA）是大氣細懸浮微粒（PM1）的主要成分之一，對於空氣品質、氣候輻射強迫以及人體健康皆具有重要的影響。本研究結合高時間解析度的實地觀測與空氣品質模式模擬，探討位於台灣台中市西屯區都市地區的有機氣膠之生成、組成與演變過程。此研究利用2023年11月觀測計畫資料，包含都市空氣汙染研究站（UAPRS）的即時mini compact time-of-flight aerosol mass spectrometer（mini-C-ToF-AMS）資料和以非即時之離子層析儀（IC）分析的濾紙採樣樣本進行的氣膠物質分析；而作為潛在二次有機氣膠（SOA）前驅物的揮發性有機氣體（VOCs）則利用台灣環境部運作的忠明光化觀測站監測資料。AMS分析結果顯示氣膠的組成與變化受通風條件與邊界層高度顯著地影響。在弱風場條件下，PM1會大量累積，且有機氣膠佔了約43%，為PM1主要成分。透過正矩陣分解法（PMF）分析，可以辨識出五種有機氣膠種類，分別為碳氫類有機氣膠（HOA）、老化碳氫類有機氣膠（aged HOA）、半揮發性氧化態有機氣膠（SVOOA）、低揮發性氧化態有機氣膠（LVOOA），以及背景有機氣膠（background OA）。此結果顯示出SOA為西屯地區之主導成分，占整體有機氣膠質量超過90%。 在模式研究使用社區多尺度空氣品質模式（CMAQ v5.3.1）結合cb6r3_ae7_aq化學機制模擬有機氣膠及其前驅物。模式結果顯示CMAQ結果與觀測之空氣污染物的整體時間變化趨勢相似，但夜間的一次有機氣膠（POA）與SOA濃度則有高估的情形。模式中POA濃度主要受排放強度與邊界層高度控制，而SOA生成則主要由人為排放源之前驅物所促進，並受大氣水平輻合條件影響其濃度的空間分布。這顯示出模式中針對排放、邊界層高度及風場的模擬誤差可能導致了POA和SOA在夜間的高估。生物源SOA主要由單萜（monoterpene）氧化生成，其主要產物包括來自單萜類有機硝酸鹽的有機假水解積聚產物（organic pseudo-hydrolysis accretion product from monoterpene organic nitrate，AMTHYD），以及乙二醛與甲基乙二醛在水相氣膠中生成的SOA（glyoxal and methylglyoxal SOA in aqueous aerosol material，AGLY）。然而，模式對土壤陽離子的高估造成氣膠酸度的中和程度偏高，進而對SOA形成的模擬造成不確定性，特別是酸催化反應途徑如異戊二烯環氧化二醇形成的SOA（acid-catalyzed isoprene epoxydiol SOA，AISO3）。本研究顯示都市地區有機氣膠生成過程的高度複雜性，並指出需進一步改善模式中針對氣膠酸度、前驅物排放量及大氣物理過程的模擬。此外，本研究也顯示高解析度觀測資料結合化學傳輸模式在掌握有機氣膠來源與生成機制方面的重要性。

Organic aerosols (OA) are a major constituent of atmospheric fine particulate matter (PM1), significantly influencing air quality, climate forcing, and human health. This study investigates the formation, composition, and evolution of OA in the Xitun urban area of Taichung, Taiwan, by integrating high-resolution field observation data with air quality modeling analysis. The analysis focuses on data collected during a field campaign in November 2023, including real-time aerosol composition measurements using a mini compact time-of-flight aerosol mass spectrometer (mini-C-ToF-AMS) and offline filter analysis via ion chromatography (IC) at Urban Air Pollution Research Station (UAPRS). Volatile organic compounds (VOCs) concentrations from Chungming Photochemical Monitoring Station, operated by the Taiwan Ministry of Environment (MOENV), were applied as potential precursors for secondary organic aerosols (SOA) formation. The results show that aerosol composition and variability were strongly influenced by ventilation and local boundary layer dynamics. During periods of stagnant conditions, PM1 accumulated substantially, with OA emerging as the dominant component. Positive matrix factorization (PMF) identified five OA components: hydrocarbon-like OA (HOA), aged HOA, semi-volatile oxygenated OA (SVOOA), low-volatility oxygenated OA (LVOOA), and background OA, with SOA accounting for over 90% of total OA mass. Simulations using the Community Multiscale Air Quality (CMAQ) v5.3.1 model with the cb6r3_ae7_aq mechanism reproduced general temporal patterns of air pollutants but overestimated both primary organic aerosols (POA) and SOA, particularly during nighttime. POA concentrations were primarily controlled by emission rates and boundary layer height, while SOA formation was driven by anthropogenic precursors and modulated by convergence patterns. Model biases in wind fields, emissions, and vertical mixing likely contribute to the nighttime overestimation. Biogenic SOA was predominantly from monoterpene oxidation, with major products including organic pseudo-hydrolysis accretion product from monoterpene organic nitrate (AMTHYD) and aqueous-phase glyoxal and methylglyoxal SOA (AGLY). However, uncertainties in modeled aerosol acidity, linked to overpredicted soil cations and excessive neutralization, likely have affected acid-catalyzed SOA pathways such as isoprene epoxydiol-derived SOA (AISO3). This study highlights the complexity of urban OA formation and the need for improved model representations of aerosol acidity, precursor emissions, and atmospheric processing. It also demonstrates the value of integrating high-resolution measurements with chemical transport models to better constrain OA sources and formation mechanisms.