探討氣候模式在全球暖化下與聖嬰現象相似之顯著熱帶太平洋海溫暖化反應

Abstract

熱帶太平洋地區的海表溫度變異可對氣候造成相當程度的影響，例如：沃克環流的強度、哈德利胞以及熱帶邊緣地區沙漠的擴張、以及熱帶降雨帶的位移。在全球暖化的情境下，大部份氣候模式認為未來熱帶太平洋地區會呈現類似聖嬰現象發生時的海表溫度變異：赤道地區加強暖化，且東赤道太平洋地區較西赤道太平洋地區暖化更為明顯。然而，由於現今氣候模式存在著氣候場以及聖嬰現象模擬的誤差，許多研究質疑此一暖化趨勢的可信度。 本研究的主要目的為探討此海表溫度暖化結構的形成機制以及評估其可信度。透過地表的能量收支分析，我們發現與現今氣候蒸發場相關且在赤道地區較弱的蒸發抑制機制、以及主要為水氣作用在赤道地區較強的向下長波輻射，為氣候模式中造成此一暖化結構的主要原因。另外，我們認為：(1)現今氣候場的誤差也可能導致未來氣候模擬的偏差，以及(2)全球暖化與聖嬰現象的形成機制並不相同，但有類似的氣候回饋機制。根據本研究的結果，透過修正現今海表溫度及能量收支的氣候場以及對聖嬰現象的模擬，能使我們對氣候模式模擬的未來海溫暖化的量值與結構更有把握。

Sea surface temperature anomaly (SSTA) in tropical Pacific region has been suggested to account for some significant climate changes, for instance, the weakening of Walker circulation, the expansion of the Hadley cell edges and subtropical deserts, and the shift of tropical rainband. Under global warming, most global climate models (GCMs) project an El Nino-like SST warming pattern, with enhanced warming in equatorial region, and a reduced zonal SST gradient along the equator. However, some previous studies have questioned the reliability of these robust responses across GCMs, since most GCMs experience difficulties simulating the observed climatology and ENSO variability in the tropical Pacific. The goal of this thesis is to evaluate the confidence of such El Nino-like SSTA pattern in tropical Pacific region among most GCMs through understanding its formation mechanisms. We demonstrate that the reduced evaporative damping related with the structure of climatological evaporation and the enhanced greenhouse effect related with increasing water vapor over the equatorial central Pacific, are the two major contributors to the SSTA pattern. Furthermore, it is suggested that (1) the biases of present day climatology in GCM simulations may lead to an error of future SST projection, and (2) the global warming scenario and El Nino events share some common atmospheric climate feedbacks, despite the distinct triggering mechanisms. The results set the foundations to evaluate our confidence of the projected SSTA in each GCM by performing a process-based comparison with the observational and reanalysis data.