熱帶環流結構變化在過去及近未來的氣候發展及預測：人為氣膠及溫室氣體彼此間的影響

Abstract

在人為氣候變遷下，哈德里環流（Hadley Circulation）將有各種方面的改變，包括間熱帶輻合區（Intertropical Convergence Zone）的位置，副熱帶沙漠的邊界，以及相關能量、水氣及動量的傳輸。在本研究中，我們藉由CESM LENS來研究哈德里環流結構變化的暫態反應演變。並透過使用「固定單一強迫項」的情境模擬量化溫室氣體和人為氣溶膠對環流改變造成的貢獻。 在 CESM LENS模擬中，從哈德里胞的結構化中可以區分出三個不同的距平環流結構的時期。第一，在20世紀後期，哈德里環流主要被人為氣溶膠主導，產生了一個順鐘向的跨赤道環流距平。這個跨赤道環流距平的發生被限制在深對流區域並可以被能量觀點的理論機制解釋。第二，在21世紀初，隨著人為氣溶膠減少以及溫室氣體增加，兩者對哈德里胞結構的改變互相競爭因此對哈德里胞的結構影響很小。第三，21世紀中葉以後，我們發現深熱帶的上升運動增強、位於南北緯5到20度的環流強度減弱，且哈德里胞的邊界擴張。而這三個由溫室氣體增加所造成的環流結構變化可被海溫結構變化解釋。 在這段兩種強迫項競爭的時期，也就是21世紀上半葉，凸顯了在當前的觀測紀錄中探測和理解強迫反應的挑戰。在2000年以後，全球均溫的上升開始變得顯著，根據對哈德里胞結構變化的理解，我們理應觀察到某些哈德里胞的結構變化：像是哈德里胞減弱、哈德里胞擴張以及深對流區的上升運動增強。然而，在模式中，我們發現這些結構變化被偵測到的時間卻因為受到氣溶膠使赤道東太平洋海溫產生冷卻趨勢的影響而被延後。也就是說，這個由氣溶膠造成的海溫冷卻趨勢在這個競爭階段相當重要。 氣溶膠對赤道太平洋海表溫度的影響可以分為快和慢兩種反應。在快速的反應中，海表溫度反應主要是透過地表海氣交互作用的過程形成，包含風—蒸發—海表溫度反饋作用（WES feedback）和風力驅動的艾克曼抽吸（Ekman pumping）。透過這些過程，在21世紀初期氣溶膠減少造成東南太平洋的冷卻，並延伸到赤道地區。在較緩慢的反應中，海表溫度的反應則和海洋副熱帶環流（Subtropical Cell）有關並造成了赤道海洋次表層的延遲冷卻。藉由這些過程，赤道海表溫度的冷卻將在2030年達到高峰，比起氣溶膠排放量達高峰的1980年代有很長時間的延遲。 藉由CESM LENS，我們了解到哈德里胞對人為強迫的反應相當複雜。溫室氣體和氣溶膠因為其各自的時間以及空間分布而在此扮演不同角色，而這需要更進一步地去探討。

Various aspects of Hadley Circulation are expected to vary under anthropogenic climate change, including the position of the intertropical convergence zone, the boundary of the subtropical desert, and the associated energy, moisture, and momentum transports. In this study, we investigate the transient evolution of the structural changes in Hadley Circulation in the Community Earth System Model Large Ensemble (CESM LENS). Using the companion “all-but-one-forcing” scenarios, the relative contribution of greenhouse gases and anthropogenic aerosol-induced circulation changes are quantified. In the CESM LENS simulations, three periods of distinct anomalous circulation structures are identified. (1) During the late 20th century, the Hadley circulation is mainly affected by anthropogenic aerosols, which causes the anomalous clockwise cross-equator cell. The cross-equator cell concentrates in the deep tropics and could be interpreted by the energetic framework. (2) During the early 21st century, as aerosols emission decreases and greenhouse gas emission increases, their effects on the Hadley Cell structures are competing and the resulted changes are small. (3) After the mid-21st century, we find the upward motion in the deep tropics strengthens, the maximums of overturning circulation in the subtropics weaken, and the poleward boundaries of the Hadley Cell expand. These three greenhouse-gas-induced structural changes could be understood via the sea surface temperature patterns. The competing period, the first half of the 21st century, highlights the challenges of detecting and understanding the forced response in the current observational records. The global mean temperature increases become significant after 2000, following the understanding of the Hadley Cell changes, the Hadley Cell structural change such as the Hadley Cell weakening, the Hadley Cell expansion, and the upward motion strengthening in the deep tropics should be seen. However, in the simulation, we find that the time of emergence of these Hadley Cell changes is delayed because of the effect of aerosols: the aerosols cause the SST cooling trend in the equatorial eastern Pacific. That is to say, this cooling trend caused by the aerosols is important in the competing period. The aerosols influence on the equatorial Pacific sea surface temperature consists of fast and slow components. For the fast component, the SST response is mainly established by the surface air-sea interaction process including the wind-evaporation-sea surface temperature feedback and the wind-driven Ekman pumping. Through these processes, the aerosols recovery at the beginning of the 21st century leads to cooling in the southeast Pacific that extends to the equatorial region. For the slow component, the sea surface temperature response is associated with changes in subtropical cells, which leads to a delayed cooling response in the equatorial subsurface. Through this process, the equatorial sea surface temperature cooling is expected to peak in 2030, much later than the peak of aerosols emissions in the 1980s. Using CESM LENS, we reveal the Hadley Cell responses to anthropogenic forcings are complex. The GHG and aerosols forcings both play a role, each consists of distinct temporal and spatial characteristics that require further investigations.