平均流場與海氣交互作用的關係─亞洲夏季季風槽之影響

Abstract

在西北太平洋夏季時(7-8月)，大氣與海洋的交互作用相當複雜，除了一般認為的大氣與海洋耦合的作用外，大氣的平均流場也扮演了重要的角色。為此，本研究特別針對西北太平洋的夏季降水，檢視其次季節(sub-seasonal)變化及年際變化，分析它們與其他變數場的關係，並利用一中型的大氣模式—QTCM測試在不同平均流場狀態下，對降雨模擬的影響。 從觀測資料顯示，在次季節變化方面：SST變化及大氣柱淨熱通量(Fnet)變化的空間分佈(pattern)與雨量的並不完全相同，且潛熱通量(LHF)在雨量上升的區域反而是下降的，因此推論雨量的次季節變化無法單靠局部區域(local)的機制，尚需大尺度的動力過程驅動。 在年際變化方面：不論哪個季節，對流區域的雨量距平大致與淨地表熱通量距平(Fsnet)呈正相關，顯示雨量的年際變化可能主要由地表熱通量距平所影響。另外，在7、8月時，雨量與SST大致為負相關，與風速則有明顯正相關，表示雨量距平並非由SST距平所驅動，而是由大氣(風場距平)來主宰。1997/98年之ENSO個案顯示，7、8月時在冷/暖SST上方、風速距平為正/負的區域，其雨量是增加/減少的，因為風速的貢獻比SST(透過水汽項)的貢獻大。 當風速距平主宰雨量變化時，平均流場的正確與否變得相當重要。利用QTCM測試平均流場的狀態—有、無季風槽對降雨模擬的影響，並分為固定雲輻射(即雲量固定)的情況與正常雲輻射(雲量會隨時間改變)的情況。檢視1998年的7、8月份，當沒有季風槽時，雨量距平主要受SST距平影響；加入季風槽後，在菲律賓附近可掌握到風速下降的趨勢，且讓負雨量距平得以西伸而位於暖SST之上，並且也使得太平洋上之反氣旋距平向西移、更接近觀測。而正常雲輻射情況下的風速及雨量距平比固定雲輻射的大一些，因此結果也較好，顯示雲輻射也扮演了一些角色。大體而言，加入季風槽後，比起沒有季風槽的情形，有向觀測接近的趨勢，顯示較好的平均流場可以改善西北太平洋夏季降雨年際變化的模擬。

Atmosphere-ocean interaction is very complex over the western North Pacific (WNP) in summer. A model with atmosphere-ocean coupling processes is one way to improve the simulation in the interannual variability over the WNP. The mean flow (monsoon trough), on the other hand, also plays an important role in simulating the interannual variation. This study first analyzed observations for possible local effects that can induce sub-seasonal and interannual variations of summer precipitation over the WNP. Then, a climate model was used to examine the influence of mean flow, monsoon trough is particular, on precipitation. The observation data show that, on a sub-seasonal time scale, the change of precipitation from June to July is not consistent with that of the net heat flux into atmospheric column (Fnet) and net surface heat flux (Fsnet), since latent heat flux (LHF) decreases over regions where precipitation increases, i.e., the WNP. Thus, the abrupt increase of the WNP precipitation from June to July is not directly contributed to by local heat fluxes. Large scale dynamics may be important for the sub-seasonal variation of rainfall over the WNP. On an interannual time scale, precipitation over convective regions is generally consistent with Fsnet. This implies that local heat fluxes may be the major factor to affect precipitation over the WNP. Observations show that the precipitation over the WNP is negatively correlated with SST in July and August, but positively correlated with low-level wind speed. In the 1997/98 El Niño, precipitation in July and August increases (decreases) over cold (warm) ocean surface where surface wind speed is enhanced (reduced), resulting in a greater contribution to LHF from wind speed than from SST. Since precipitation anomaly is more associated with wind speed anomaly in the interannual time scale over the WNP, how good the mean flow is simulated in climate models becomes critical. Model experiments indicate that surface wind speed and precipitation anomalies are closer to observations when the summer monsoon trough is prescribed. In other words, the mean state of the WNP monsoon trough can affect the interannual variation of precipitation over the WNP.