| 摘要: | 大氣邊界層內大氣垂直結構的日夜變化會影響低層大氣的空氣污染物濃度。但傳統大氣探空觀測每日僅進行兩次，不足以解析大氣垂直結構隨時間的變化；此外傳統大氣探空觀測項目僅限於氣象參數，無法知道垂直方向污染物濃度的變化。因此本研究嘗試開發一款微型大氣探空系統Aerosond，因成本低於傳統探空，故可更密集地施放以取得較高時間解析度的大氣資料，且Aerosond探空儀裝有PM感測器，可提供垂直PM2.5濃度資料。
在2021年至2023年間，Aerosond經改良溫濕度探頭的設計與升級gps天線等零件後共迭代4個版本，每次版本更新皆與氣象署於板橋站施放的Vaisala探空儀進行平行比對實驗，驗證其準確度與效能。平行比對結果顯示最終版本在3 km以下溫度的平均絕對誤差為0.6°C，其中夜間溫度的平均絕對誤差僅有0.2°C；3 km以下相對濕度、風速、風向之平均絕對誤差分別為4.5%、0.3 m s-1、7°。
每年冬末春初適逢臺灣中南部空污季節，Aerosond參與國科會臺灣大氣邊界層觀測實驗(T-POMDA)，於2021至2023年間共進行四次密集觀測，每次以Aerosond探空進行一至四點的同步觀測。垂直剖面資料分析結果指出，造成空污的原因除風速弱導致水平擴散條件不良外，還包括較強的輻射逆溫強度和較高的大氣穩定度使垂直擴散條件不佳。此外沿海地區與內陸地區在空污事件下的垂直結構有所不同：沿海地區可能全日皆有逆溫層存在，使近地面PM2.5全日皆維持高值；內陸地區白天穩定度較低，PM2.5可垂直擴散到比沿海地區更高的高度，但內陸地區夜間輻射逆溫強度傾向強於沿海地區，使內陸地區夜間PM2.5累積於低層。利用Aerosond資料測定邊界層高度，本研究發現在空污事件期間使用bulk Richardson number法可獲得較合理之邊界層高度。值得一提的是，由於Aerosond探空可涵蓋10 km以下的觀測，所以可用其資料確認東南亞生質燃燒跨境傳輸PM2.5污染物分布的高度與濃度。
從平行比對實驗與T-POMDA空污實驗的觀測成果說明，Aerosond微型探空之資料具備相當的準確度，有助於解析大氣結構對PM2.5垂直濃度分布之影響，有助於探討高空污事件的發生成因，並可以做為後續模式模擬驗證或資料同化等應用。;The structure of planetary boundary layer (PBL) changes in a day and influences the concentration of low-altitude pollutants. However, the traditional routine radiosonde observation is conducted only twice a day and can not resolve the diurnal variation of the atmosphere. In recent years, miniature radiosonde systems were developed to meet the need of intensive observation. Thus, we developed a miniature radiosonde system with measurement of PM2.5 in this research, and name it Aerosond.
We improved the design of temperature and relative humidity sensors of Aerosond and upgraded its GPS sensor, to release four versions of Aerosond during 2021-2023. Every version of Aerosond conducted the data quality validation experiments with Vaisala RS-41 radiosonde. The results of data quality validation experiments showed that under 3 km, the latest version of Aerosond has a mean absolute error of temperature with 0.6°C (nighttime temperature of 0.2°C), relative humidity with 4.5%, wind speed with 0.3 m s-1, and wind direction with 7°.
Because weak synoptic weather made air quality worse in Central and Southern Taiwan in winter and spring, T-POMDA experiment was held to study the interaction between meteorology and air pollution. Aerosond were applied in T-POMDA experiment to have intensive observation under four air pollution events in central and southern Taiwan during 2021-2023. Vertical profile data indicate that the meteorological factors which lead air pollutants accumulated within PBL including low horizontal wind speeds, stronger radiative cooling intensity, and high atmospheric stability. Coastal regions and inland regions have different characteristics under air pollution events. In coastal regions, inversion layer may exist the whole day and lead to high PM2.5 concentration at near-surface. In daytime of inland regions, the atmospheric stability is lower than coastal regions and let PM2.5 disperse to higher altitude. Whereas in nighttime of inland region, there has strong inversion intensity and make PM2.5 accumulate in low altitude. During intensive observation period, bulk Richardson number method can provide us reasonable PBL height. It’s also worth mentioning that Aerosond’s data can be used to identify the height and concentration of long-range transported biomass burning PM2.5 from Peninsular Southeast Asia in free atmosphere, because Aerosond can cover the observation under 10 km.
The results of data quality validation experiments and the results of T-POMDA experiments indicates that Aerosond can provide reliable data, and can help us to study how atmospheric structure affects the distribution of PM2.5. Aerosond’s data can validate numerical model or be applied in data assimilation in the future. |
