查詢結果分析
來源資料
頁籤選單縮合
題 名 | 衛星觀測輻射資料之一維變分反演法=One Dimensional Variational Retrieval Scheme on the Satellite Observed Radiance |
---|---|
作 者 | 周鑑本; 李尚武; | 書刊名 | 大氣科學 |
卷 期 | 25:3 1997.09[民86.09] |
頁 次 | 頁337-356 |
分類號 | 328.886 |
關鍵詞 | 反演; 資料同化; Retrieval; Data assimilation; |
語 文 | 中文(Chinese) |
中文摘要 | 本實驗利用一維變分及疊代法進行數值模式直接使用衛星觀測資料的測試。 實驗 首先以理想化大氣作為校驗用的〞真值〞,然後在此理想化大氣與由輻射模式計算出此理想 化大氣下的輻射值中分別加入隨機誤差,模擬數值預報模式所提供的背景剖面與衛星觀測值 。實驗亦選取實際資料進行初步的反演測試。理想資料測試結果顯示,此變分反演法可以依 衛星觀測來修正背景剖面而減少其誤差。實驗的敏感度測試顯示,背景場參數的雲高誤差主 要影響反演的收歛個數, 但是並不影響對溫溼剖面的調整。 而且當雲高的誤差標準差在 100hPa 與 50hPa 時,雲高與雲量則有明顯的修正。反演結果的另一主要現象是背景場誤差 的垂直相關性與誤差的修正量間約成正比。對於溫度剖面的修正量而言,中高對流層的誤差 約減少 0.5 ° K 至 1.0 ° K。而水氣( g/Kg )修正量的對數值約可減少 0.1 的誤差, 主要亦發生在中高對流層。從各項測試中的發現地表的溫度有 0.6 ° K 至 1.0 ° K 的修 正量,但是對地面空氣溫度則無影響。 在尚未完成實際背景場的誤差統計前,對實際 TOVS 衛星觀測資料個案的反演測試採用假設的誤差協方差矩陣,反演結果仍有較小量的誤差修正 。我們推估此不明顯的誤差修正量主要的原因為假設的協方差矩陣與實際的背景場誤差性質 有異所致。因此,若以數值模式預報為背景場則模式的誤差統計是改進反演品質的重要因素 之一。 |
英文摘要 | The motivation of this study is to achieve the direct use of the satellite observed radiance in a numerical weather prediction model. This experiment utilizes the one-dimensional variational (1D var) analysis and an iteration scheme to assimilate the radiance of the idealized atmospheric profiles and the observed TOVS radiance. The idealized profiles are used as a "true value" for the evaluation in this experiment. Random errors are added to the idealized data for the background and simulated satellite observations. The preliminary results of this experiment show the variational assimilation scheme can reduce the analysis error from its background error. The temperature analysis in the middle and upper troposphere has been improved 0.5 ° K to 1.0 ° K from its background error. The mixing ratio (g/Kg) in logarithm has also shown a 0.1 improvement in the similar level. The on the temperature and moisture improvements are not sensitive to the error of the initial cloud height estimatin. However, the initial error of the cloud height affects the number of the convergence in the iterational scheme. Significant improvements of 0.6 ° K to 1.0 ° K are found on the soil temperature. The error covariance plays an important role in the variational assimilation scheme. It receives minor improvements by applying the hypothetical error covariance matrix to assimilate the real TOVS radiance. This may due to the error covariance matrix is not consistent with the characteristics of the numerical model forecast. |
本系統中英文摘要資訊取自各篇刊載內容。