GMS地球同步衛星影像數位資料在雲分析上的應用

The Application of GMS Digital Image Data to Cloud Analysis

林博雄(Po-Hsieng Lin) ； 曾忠一(Chung-Yi Tseng)

《大氣科學》 22卷3期 (1994/09) Pp. 319-337

摘要

傳統主觀氣象雲圖分析方法有視覺限制與錯覺，定量與自動化的雲圖分析，將可有效地組合雲特徵與數值天氣預報的產品。本文嘗試建立一套資料處理流程，以1990年6月12日0000GMT為例，結合日本GMS地球同步衛星雲圖數位資料與歐洲中期天氣預報中心（ECMWF）數值天氣網格資料，分析東亞地區梅雨鋒面雲帶、赤道對流雲簇等豐富的雲系現象。本文中首先探討紅外雲圖中臨邊昏暗現象的視角訂正問題，在LOWTRAN程式套的輔助下，求得（secθ）^(1/8)的經驗式。經過視角訂正後的亮度溫度與反照率，再配合數值模式網格，以大約150公里見方的空間網格進行雲量、雲頂溫度與反照率平均值的計算，並以空間相干法另外再計算衛星資料網格內有雲覆蓋區的雲頂溫度。接著以方格區分法和日本氣象衛星中心的雲級分類技術，組合雲頂溫度與反照率值來討論雲型與雲級的空間分布。這些雲型、雲級相當程度地掌握到縱觀尺度雲系分布的特徵，並和ECMWF數值模式的垂直速度場有相當程度的吻合，但與對流可用位能（CAPE）的大小沒有良好的對應關係。

關鍵字

地球同步衛星 ； 雲分析 ； 雲參數遙測

並列摘要

The traditional, visual method for satellite cloud map analysis has the problems in judgemental error and vision constraint by human eyes. Quantitative and automatical analysis of satellite cloud map can help to combine numerical weather prediction products and cloud information more efficiently. This paper attempts to outline a data analysis procedure, which use the GMS digital data and TOGA advanced grid data from European Centre of Medium-range Weather Forecasts. The case was chosen on 0000GMT 12 June, 1990, in East Asia area. There are Mei-Yu (Baiu) frontal system, cumulus convections and cloud clusters near the equator. We first discuss the limb darkening phenomenon in infrared band and the correction algorithm. By using LOWTRAN package the correction function is found to be (secθ)^(1/8). After the correction stage reflectivity and brightness temperature were remapped into about 150km square to match the ECMWF numerical dataset. Then the grid average of cloud amount, cloud top temperature and cloud reflectivity were retrieved. We also follow the spatial coherence method to compute cloud top temperature in the satellite grid area. Finally, the quantitative analyses of GMS satellite visible and infrared digital data were done in two paths. One approach is using box classifer idea which uses several intervals of cloud top temperature and reflectivity. Five cloud patterns were assigned as five interval boxes. The second approach is the same as the idea of Japan Meteorological Satellite Center. Three degrees of cloud top height were seperated and identified in the space. The spatial distribution of these cloud patterns and cloud height degree catch most synoptic features. The comparison between these cloud analyses and ECMWF vertical velocity field has a good agreement but the convective available potential energy (CAPE) variable fails to fit these cloud analyses position.

並列關鍵字

Geosynchronous satellite ； Nephanalysis ； Remote sensing of cloud parameters