| 摘要: | 汞(Hg)是一種具有生物累積性的重金屬，能夠經由大氣傳輸影響全球，因此大氣的種種機制都能影響大氣汞傳輸。以往鹿林山大氣汞的研究多著重在周期較短的變化，如日變化、月變化或季變化等，而周期較長的年際變化則較少被研究。因此本研究使用希爾伯特-黃轉換(Hilbert-Huang transfer)來分析2006年到2022年鹿林山大氣汞的年際變化，並分析主要的影響因素。
2006到2022年鹿林山氣態元素汞(Gaseous elemental mercury, GEM)平均濃度為1.55ng/m3，標準差為0.37 ng/m3，使用Mann-Kandall test及Sen’s slope method計算後發現，這17年間GEM濃度有-0.022ng˙m-3•year-1的遞減趨勢，此趨勢可能與排放源的排放量減少有關。在使用希爾伯特-黃轉換分析GEM資料後，得到12個本質模態函數(Intrinsic Mode Function, IMF)，其中IMF1到IMF8是屬於日變化、月變化與年變化的部分，主要是由於日夜轉變、山谷風、不同季節的污染物傳輸差異等所造成。IMF9到IMF11則是屬於年際變化的部分，將其與南方震盪指數(Southern Oscillation Index, SOI)的週期相近的IMF進行分析及比較後發現，在聖嬰現象期間鹿林山的大氣汞濃度較高，反聖嬰期間的大氣汞濃度較低。此外，由於春季中南半島生質燃燒是鹿林山重要的大氣汞來源，因此本研究將美年三月的資料串連起來進行分析，發現其的確有與聖嬰現象相似周期的IMF，並且在總變異量中占了不小比例。
此外，在分析過每年三月的月平均大氣汞及nino3.4資料後，發現兩者的相關性良好(r=0.48)，能夠佐證大氣汞濃度與聖嬰現象的關係。進一步計算聖嬰年與反聖嬰年的氣流距平場後，發現在聖嬰年時氣流會有更多由西南向東北的分量，因此帶來更多春季東南亞生質燃燒產生的的污染物，導致大氣汞濃度的上升；反聖嬰年時則有更多由東向西的分量，帶來更多海洋的氣團使得大氣汞濃度較低。此外，2013及2017這兩年並非聖嬰或反聖嬰年卻也有大氣汞濃度的極值，也是由於氣流方向的不同所導致的。
另外，本研究也分析了其他因子對春季大氣汞濃度年際變化的影響，包括東南亞的火點數量、溫度、輻射量等因子，而這幾項因子與大氣汞濃度的相關性都不佳，因此推論聖嬰現象是影響大氣汞年際變化的重要因子。
;Mercury (Hg) is a bioaccumulative heavy metal that can affect the world by atmospheric circulation. Therefore, various atmospheric mechanisms can affect the transport of atmospheric mercury. Previous studies at Lulin Atmospheric Background Station (LABS) focused on short-period variabilities such as diurnal variability, monthly variability and annual variability. However, interannual variability have been less studied. Therefore, this study will use Hilbert-Huang transform to analyze the interannual variability at LABS from 2006 to 2022, and find its main influencing factors.
The average concentration of gaseous elemental mercury (GEM) at LABS from 2006 to 2022 was 1.55ng/m3, and the standard deviation was 0.37 ng/m3. After using the Mann-Kandall test and Sen′s slope method, it was found that there were a downward trend of -0.022ng˙m-3·year-1, which may be related to the reduction of emissions from emission sources. After analyzing the GEM data by Hilbert-Huang transform, it was found that 12 intrinsic mode functions (IMF) can be obtained. The IMF1 to IMF8 are diurnal to monthly to annual variabilities, which is mainly caused by diurnal changes, valley winds, differences in pollutant transport in different seasons, etc. IMF9 to IMF11 are part of the interannual variabilities, and we comparing them with the cycle of the Southern Oscillation Index (SOI). After analyzing and comparing similar IMFs, it was found that there is a phase difference between the two in time. Therefore, when the El Nino phenomenon occurs, the atmospheric mercury concentration in Lulin Mountain will be higher, and vice versa. In addition, since biomass burning in spring is an important source of GEM at LABS, this study also connected the data in March for analysis and found that it does have an IMF with a similar period to the El Niño-Southern Oscillation (ENSO), and the total variation is higher accounted for a large proportion.
After analyzing the monthly average GEM concentrations and nino3.4 data in March every year, it was found that the correlation between the two is good (r=0.48), which can support the relationship between atmospheric mercury concentration and the ENSO. And further print the streamline anomaly of El Niño and La Niña years on March, we found that there was more airmass from the southwest to northeast in El Niño years, which would bring more air pollutants produced by biomass burning in Southeast Asia and caused higher GEM concentrations. In contrast, more airmass from the east in La Niña years, which brought more oceanic and cleaner airmass and made GEM concentrations at LABS lower. In addition, the two years 2013 and 2017 were neural years, but they also had extreme values of GEM concentration, which was also caused by the different air flow directions.
This study also analyzed the impact of other factors on the interannual variability of GEM in spring, including the number of fire points in Southeast Asia, temperature, radiation and other factors that may affect the GEM concentration. But the correlation is not good, so it can be inferred that the ENSO is an important factor affecting the interannual variability of atmospheric mercury. |
