因應氣候變遷情境下，全球天氣與氣候模式的開發刻不容緩，因此模式必須全方位地進行診斷及評估，方能切合各方所需，其中衛星資料不僅具有高空間及高時間的觀測優勢，亦可提供輻射總量（地表、大氣和雲）之觀測訊息，如結合輻射傳送物理模式之計算，對於數值模式在雲、水氣與輻射等物理量預估結果之診斷與模式物理過程之修正，可提供相當客觀的參考，為極佳的模式評估應用工具之一。
然而衛星提供的觀測資料與數值天氣預報模式輸出之參數是在不同的物理空間上，衛星觀測為輻射量，而數值模式輸出為地球物理量(如溫度、濕度等)，因此我們無法將兩者直接進行比較。在本研究中使用中央氣象局全球預報模式輸出作為診斷案例，透過輻射傳輸模組，模擬晴空狀況下的水氣頻道與窗區頻道之亮溫(即6.7 and 11μm，並使用歐洲中期天氣預報中心之在分析資料作為診斷參考。
本研究診斷了2015年5月的全球表面溫度以及中高層(400至600hPa)之濕度場，初步結果顯示兩模式間全球水氣總量相當一致，但在表面溫度在高緯度地區，以及乾燥地區有明顯的差異，並且兩模式間的日夜變化強度在乾燥地區表現也不甚一致。
;Numerical weather prediction (NWP) models are the essential tools that provide short-term (a few seconds) weather forecasts to longer-term (larger than decades) climate predictions. It is essential to evaluate and quantify the NWP model forecast capability so that the improvement can be conducted through several techniques such the using of either parameterization scheme or initial/boundary conditions.
Satellites are very useful tools for diagnosing the performance NWP models because they provide near-global coverage and high temporal resolution data. However, the satellite observations and the NWP models’ output are in difference space (e.g., radiative quantity and geophysical fields). Therefore, these two spaces data sets cannot be compared directly. In this study, we select Taiwan Central Weather Bureau global forecast system (CWBGFS) as example. Using radiative transfer model (RTM) to simulate the satellite nadir-like observation at the wavelengths of 6.7 and 11μm, which represent for upper tropospheric water vapor and surface temperature, respectively. Similar RTM calculation are carried by adopting the European Centre for Medium-Range Weather Forecasts (ECMWF) re-analysis interim (ERA-interim) data set as reference.
The preliminary results demonstrate early assessment results in the month of May in 2015, which is the Asia monsoon season. The polar region reveals a relative high surface temperature, and the diurnal variability is not significant in CWBGFS.
