本研究使用歷年環保署空氣品質監測站數據，分析高屏地區臭氧及其前趨物濃度的周變化型態，發現高屏地區與國外許多都市中心一樣，周末光化反應前趨物濃度降至最低，但臭氧濃度反而比平日高的＂周末效應＂現象；而且發現臭氧濃度周變化有逐日累積的趨勢，並在周六到達最大值。 我們發現一些現象有利於臭氧周末效應，首先，觀察臭氧、Ox 和NOx 濃度周變化，氮氧化物濃度在周末減少會降低NOx 滴定效果，因此在周末有利臭氧濃度增加；其次，PM10 濃度均周末低於平日，這可能會增加太陽輻射到達地面，因此有利光化學反應；高屏地區都市下午O3/NOy 中位數皆落在5 左右，可概略判定為VOC-sensitive，周末NOx 排放量降低有利於臭氧濃度增加。 我們也發現前日午後總氧化劑濃度與當日最大八小時臭氧平均濃度呈現正相關，此可解釋臭氧濃度累積的現象；而當日上午NO 濃度與當日臭氧濃度呈現負相關，可能與NOx 滴定效應有關；我們用前日午後總氧化劑濃度和當日上午NO 濃度作自變數，建立一線性回歸模式以推估當日最大八小時臭氧平均濃度，所求得之臭氧濃度估計值與實測值的相關係數平方為0.3，証實此兩項因素能解釋臭氧濃度變化的30％。 The observations obtained from the air quality monitoring stations operated by Taiwan EPA were used to determine the weekly patterns of ozone concentrations in Kaohsiung and Pingtong area. It is found that the weekend concentrations of ozone precursors were lower than that of the weekdays; on the other hand, the weekend ozone concentrations are higher than that of weekdays. Similar to many large cities in the world, the so called “weekend effects of air quality” also occurs in this area. Moreover, it is noted that the daily averaged ozone concentrations were increased monotonically from Monday to Saturday. Several evidences were found that can explain the occurrence of the weekend effect. Firstly, we used the weekly patterns of O3, Ox, and NOx to show that the reductions of NOx concentrations on weekend may decrease the NOx titration effect. Hence, the weekend O3 concentrations will become higher. Secondly, the concentrations of PM10 were lower on weekend, which may increase the solar radiation that reaches the ground and thus increase the photochemical reaction rates. The median of afternoon O3/NOy were about 5, which indicates that this area is VOC-sensitive. The reductions of emission rates of NOx on weekend may increase the O3 concentrations. We assume that the Ox may store in the elevated residual layer at night, and mixed with air near the ground next morning to increase the ground-level O3 concentrations. This assumption is confirmed by the fact that the previous days’ afternoon Ox concentrations were correlated with the maximum 8-hr O3 concentrations. A linear regression model was built to estimate the maximum 8-hr O3 concentration from the previous day’s afternoon Ox concentrations and morning NO concentrations. This model can explain 30% of the variations of the maximum 8-hr O3 concentrations.
