東亞夏季季風期間雲輻射效應影響低頻振盪的模擬研究

Simulation Studies of Cloud/Radiation Effects on Low Frequency Oscillation during the East Asian Summer Monsoon Season

柯文雄(Wen-Shung Kau) ； 吳奇樺(Chi-Hua Wu) ； 鄒治華(Chih-Hua Tsou) ； 杜嘉修(Chia-Hsiu Tu)

《大氣科學》 31卷2期 (2003/06) Pp. 159-180

摘要

研究大氣低頻振盪的重要目的，就是要將其結果用於長期天氣和短期氣候預報，大氣中季內值的大小與大氣輻射有直接的關係，所以要想對氣候研究找到較好的代表性，則需同時在對雲的水平雲量，垂直方向雲的分布、雲高、雲的光學性質有好的掌握。若數值模式能模擬季內振燙的強度，那麼也必能模擬較好的季風降雨，目前許多大氣環流模式並不能正確的模擬出亞洲季風的平均結構及其季內振盪的強度，可能與無法正確的模擬出此地區雲與輻射的特性有關。我們認為影響OLR及其季內振盪的強度最重要的因子就是雲的水平及垂直結構。透過ERBE與ISCCP-DI及D2的資料我們驗證了東亞夏季季風期間，強深對流區(孟加拉灣、南海及西北太平洋)的OLR、高雲及雨量等三者有非常密切的關係，再透過濾波20-80天季內振盪的分析也得到OLR與高雲雲量的振幅變化是一致的。因而GCM對東亞夏季季風期間深對流區域的模擬，若高雲的模擬不正確則LOR就會有差錯，高雲的低頻擾動模擬的太小則LOR的擾動也會跟著降底。透過NTU-GCM的實驗模擬我們得到了控制實驗大致能模擬東亞夏季季風期間LOR、高雲與雨量在深對流區的時空平均分佈，而且對季內振盪的變化亦可掌握。
另外由三個不同雲量設定的實驗模擬我們可以得到：(1)每個模擬對東亞夏靜季風的動力場與雨量的平均狀態皆大同小異，且與其觀測的氣候平均差異不大，(2)由於雲量設定的不同會造成OLR的劇烈變化，控制實驗在孟加拉灣、南海及西北太平洋皆為底OLR值區(低於210W/M2以下)，但其餘三個實驗的值在此三個區域皆偏大(大於控制實驗50W/M2以上)，尤其沒有雲量的模擬其OLR值為280(W/M2)以上，(3)從OLR總異量的比對中圖以得到除了控制實驗的模擬與觀測分析較接近外，其餘的三個實驗的總變異量皆非常小且無雲及固定月雲量實驗模擬的空間的分佈也不正確，(4)從20-80天變異量的比對中亦可得到如(3)的狀況，(5)由於固定月雲量實驗的雲量非由模式本身所產生的，因而所模擬的OLR較其他實驗較不一樣，進而影響到其他氣象場的分佈，再者因雲量變化的時間尺度盥月，故OLR等的低頻變化非常小，(6)沒有中雲的實驗與控制實驗的結果較接近，但由於慮中雲故模擬的OLR值皆較控制實驗大，同時亦造成底頻的變異大大減小，顯示出中雲的存在是必要的。有關雲的水平及垂直分佈的分析研究對了解東亞夏季季風是刻不容緩的，研究只是初步的起始。

關鍵字

東亞夏季季風 ； 雲/輻射效應 ； 低頻振盪 ； 全球環流模式 ； LOR

並列摘要

The study of low frequency oscillations not only attempts to improve the forecast of long-term weather patterns and short-term climate change but also provides a better theoretical understanding when using numerical weather prediction guides. In general, the analysis of intra-seasonal oscillation (ISO) is based on LOR data and the magnitude of such data is strongly correlated with cloud structures in the atmosphere. Therefore, we can’t ignore the detailed structure of clouds’ vertical and horizontal distributions to gain a better representation of ISO. Currently, many models used by the atmospheric general circulation model (GCM) do not simulate the mean structure of the Asian summer monsoon with sufficient accuracy and especially ISO and could be due to accurately simulate cloud formations and the radiating properties of clouds. Analysis of the ERBE,ISCCP D1and D2data sets have revealed a strong correlation between high cloud amount, the OLR, precipitation and dense areas of convection in the Bay of Bengal, South China Sea and Western North Pacific during the Asian summer monsoon season. Using the 20-80 day filter method we can also find OLR and high cloud amount have similar phase structure and suggests that if we want to have a better simulation of ISO then the evolution of high clouds and cloud density are vital to improve the GCM.
In controlled experiment using the NTU-GCM, we can simulate the temporal and spatial patterns of OLR, high cloud amount, precipitation and ISO with reasonable accuracy. Based upon an other three carefully experiments using different settings of cloud amounts, we obtained the following results:
(1) Each experiment displays the same spatial patterns of wind field and precipitation when averaged over the whole summer and compared to observed climate averages.
(2) In the control run, the value of OLR in areas exhibiting deep convection can be lower than 210 W/m2 The other three experiments the OLR display a much larger variance than in the control run with values exceeding 280 W/m2.
(3) The total variance of OLR in the simulated control run is in good agreement with recorded observations. The other three experiments differ greatly from control experiment, not only in magnitude but also in spatial distribution.
(4) Using the 20-80 day variance analysis for the four experiments, we can obtain the same results found in our third conclusion (3).
(5) In the fixed cloud amount experiment, where the cloud amount is determined by the observed monthly mean data, the simulated OLR distribution is different from the other experiments.
(6) In the no middle cloud experiment, the simulated result is very similar to the control run. As might be imagined, its OLR value is much higher than the control run and its OLR variance is greatly reduced and demonstrates the importance of middle cloud effects upon the magnitude of OLR in a deep convection system.
Cleary, the role of clouds played in the evolution of the East Asian summer monsoon season need more detailed study, in particular in the aspects of heating/cooling attributed to clouds. This study represents a starting point in the research of the radiating effects of clouds on an existing climate system

並列關鍵字

East Asian summer Monsoon ； Cloud/radiation effect,Low frequency Oscillation ； GCM ； OLR