2018年春季鹿林山測站主要有山谷風和生質燃燒事件，相較於非事件PM2.5質量濃度的10.2 ± 6.8 μg m-3，生質燃燒事件和山谷風事件的PM2.5質量濃度分別上升至21.8 ± 6.6和20.1 ± 1.3 μg m-3。PM2.5水溶性無機離子在非事件、生質燃燒事件、山谷風事件平均濃度別為4.9 ± 3.1 μg m-3、5.7 ± 2.0 μg m-3、9.9 ± 0.7 μg m-3，顯然污染事件帶來較多的PM2.5水溶性無機離子。在豐原高中和中央大學5月高濃度事件中，PM2.5質量濃度分別為23.7 ± 6.8和21.2 ± 2.6 μg m-3，在中央大學4月的PM10質量濃度事件則達80.6 ± 5.5 μg m-3，在中央大學和豐原高中量測期間PM2.5/PM10分別約為0.3至0.6以及0.2至0.4，顯示這兩個地方粗粒徑微粒在PM10占有較大比例。比較NO3-和SO42-在微粒中占比，在PM10中NO3-大於SO42-，但在PM2.5中卻是SO42-占比較大，表示在PM10中有很多的粗粒徑NO3-。在平地監測到的NO2-多在夜間生成，且與相對濕度以及NO2濃度有較大的相關性。鹿林山的PM2.5事件主要受到生質燃燒煙團長程傳輸以及人為傳輸污染影響，在中央大學以及豐原高中的高微粒濃度普遍受到海陸風影響，特別是在夜間常因風速和邊界較低而導致污染累積。
本文發現在低相對濕度、非靜風、無雲霧的環境條件下，鹿林山大氣氣膠化學特性和山頂以上的大氣氣膠光學厚度(Aerosol Optical Depth, AOD)相關性良好(R2=0.68，p< 0.05)；即使在平地地區，非靜風的環境下，大氣氣膠化學特性和整層大氣氣膠光學厚度也具有關聯性(R2=0.5，p< 0.05)。本文使用ISORROPIA Ⅱ 模式進行氣膠酸度模擬，顯示在高山主要為酸性氣膠而平地多為酸性或是中性氣膠，利用各離子莫耳濃度計算的相關性對高山或是平地氣膠水溶性無機離子結合型態做推估，顯示鹿林山在生質燃燒事件有較多的硫酸鉀、硝酸鉀，平地在微粒高濃度事件則以硝酸銨為主。
;Water-soluble inorganic ions (WSIIs) of atmospheric aerosol have a significant effect on the atmospheric environment. These inorganic ions need to observe with high time-resolution as they change their properties rapidly in the environment. This study measured PM2.5 WSIIs with the semi-automated method at the Fengyuan High School (FHS) in November 2017, Lulin Atmospheric Background Station (LABS) in March-April 2018, and at National Central University (NCU) in April-May 2018. The measurements were toward PM2.5 most of the time; however, for some time, the system switched to PM10 WSIIs. The results accompanying related monitoring data at the stations were suitable for investigating aerosol chemistry, physics, optical properties, and source contributions.
The averages of PM2.5 mass concentrations for the events of biomass burning (BB) and mountain-valley (M-V) wind were 21.8 ± 6.6 and 20.1 ± 1.3 μg m-3, respectively, in contrast to that of non-event PM2.5 mass concentration of 10.2 ± 6.8 μg m-3. Meanwhile, the averages of PM2.5 WSIIs at non-event period, BB, and M-V wind events were 4.9 ± 3.1 μg m-3, 5.7 ± 2.0 μg m-3, and 9.9 ± 0.7 μg m-3, respectively. Obviously, pollution events brought more PM2.5 WSIIs from the comparison. In high concentration events at the FHS and NCU (May), the averages of PM2.5 mass concentrations were 23.7 ± 6.8 and 21.2 ± 2.6 μg m-3, respectively. In contrast, the high concentration event at NCU (April) was as high as 80.6 ± 5.5 μg m-3 in PM10. The PM2.5/PM10 ratios were 0.3-0.6 and 0.2-0.4 at NCU and FHS, respectively. Evidently, coarse particles were more in PM10 at both sites. Comparing the fraction of NO3- and SO42- in particles, NO3- is higher than SO42- in PM10, but the reverse is true in PM2.5. It implies that a lot of coarse NO3- existing in PM10. For monitoring results in the ground level, NO2- observed to form during nighttime and correlated well with relative humidity (RH) and NO2 concentration. The PM2.5 events were under the influences of long-range transport of BB smoke and anthropogenic pollution transports at LABS. In contrast, the land-sea breeze was mainly responsible for high particulate concentrations at NCU and FHS, especially for pollution accumulation from low wind speed and shallow boundary layer at night.
As revealed from this study, low RH, non-calm wind, and non-fog environment provided a good correlation between aerosol chemical properties and aerosol optical depth (AOD) at the summit of LABS (R2=0.7, p< 0.05). Even in the ground level, aerosol chemical properties and AOD correlated moderately with each other under a non-calm wind environment (R2=0.5, p< 0.05). The application of the ISORROPIA II model for aerosol acidity simulation indicated that mountain aerosol was acidic, and the aerosol in the ground level was more toward acidic or neutral. The computations of molar concentrations of the correlated WSIIs to infer aerosol compound form resulting in more potassium sulfate and potassium nitrate in BB events at LABS and ammonium nitrate in high particulate events in the plain area