就氣象上而言，校驗的最終目的是要增進天氣預報能力，數值模式的校驗則是要能充分掌握模式的特性，以期於使用模式時，根據模式表現特性，適當修正其誤差。本研究是利用美國國家環境預報中心（NCEP）發展的舊版（f97）與新版（NRSM）有限區域波譜模式，並選取「東北季風時期」與「梅雨季時期」個案進行模擬，RSMA部分以針對大尺度系統之掌握及預報為主，MSM部分則以觀測站資料為參考值，校驗模式的準確度與誤差。
由研究分析顯示：f97與NRSM於「東北季風時期」RSMA部分，地面的氣壓場、溫度場、風場及低層的高度場、溼度場均呈現NRSM之結果比f97較佳；而MSM部分，氣壓場以NRSM較接近實測值，溫度場部分NRSM呈現異常下降現象，f97無此現象且較接近實測值，風場結果則顯示兩模式差異較小。而「梅雨季時期」RSMA部分，地面的氣壓場、溫度場、風場及低層的高度場、溼度場均呈現NRSM之結果比f97較佳；而MSM部分，氣壓場與溫度場以NRSM較接近實測值，且無「東北季風時期」溫度異常下降現象，風場結果則仍顯示兩模式差異較小。
本研究結果顯示，新版NRSM模擬結果整體比f97好，顯示新版本的區域波譜模式有相當程度的改良與精進，且藉由與分析場校驗與比較之結果，可讓使用此模式之作業單位，掌握與瞭解模式預報的特性，並藉此特性調整預報產品，且可進一步調整模式設定來改善預報能力，達到精進區域波譜模式之目的。

On the meteorological point of view, the ultimate goal of verification is to improve weather forecasting ability and model verification is to take advantage of the characteristics of models completely, in order to amend the error. This research utilizes the earlier (f97) and new versions (NRSM) of the nested regional spectral model developed by the U.S. National Centers for Environmental Prediction (NCEP). This research selects the "northeast monsoon period" and "Mei-Yu season period" case simulation, RSMA is to grasp the large-scale systems and forecast, MSM part is to verify the accuracy of model with the observatory data, checking the accuracy and error of model.
By the analysis showed that f97 and NRSM in the "northeast monsoon period" RSMA, surface pressure, temperature, wind and lower height field and humidity field results are presented NRSM better than f97. For the MSM of NRSM, surface pressure field closer to measured values, temperature dropped abnormally, but for MSM of f97 no such phenomenon and closer to the measured value. The wind field, however, shows two models disparity. The " Mei-Yu Season period " RSMA, surface pressure, temperature, wind and lower height field and humidity field results are presented NRSM better than f97; For the MSM of NRSM, surface pressure field closer to measured values, and no " temperature dropped abnormally", the wind field results show that the two models is still little difference.
The results show that the NRSM is better than f97 overall, show that the NRSM with a considerable degree of improvement. By the results of comparing with the analysis field, it can make the operation units of using this mode to master and understand the characteristics of the model forecast, and to adjust the properties, and may further adjust the mode set to improve forecasting capabilities, to achieve the purpose of model improvement.

誌謝 ii
摘要 iii
ABSTRACT iv
目錄 v
表目錄 vii
圖目錄 viii
1. 前言 1
1.1 研究動機 1
1.2 研究目的 1
1.3 論文結構 2
2. 文獻探討 3
3. 模式介紹 5
3.1 模式預報方程 6
3.2 模式的物理過程 7
3.3 模式的數值方法 8
3.4 模式的格點結構 8
4. 研究方法與步驟 9
4.1 模式設定 9
4.2 資料來源與個案選取 9
4.3 研究方法與步驟 10
4.3.1 研究方法 12
4.3.2 研究步驟 12
5. 模式預報資料分析 15
5.1 東北季風期間模式模擬分析 15
5.1.1 個案一綜觀天氣概述 15
5.1.2 個案二綜觀天氣概述 15
5.1.3 靜力模組預報差異分析 16
5.1.4 非靜力模組預報差異分析 23
5.2 梅雨季期間模式模擬分析 25
5.2.1 個案一綜觀天氣概述 25
5.2.2 個案二綜觀天氣概述 25
5.2.3 靜力模組預報差異分析 25
5.2.4 非靜力模組預報差異分析 30
5.3 綜合分析與比較 31
6. 結論與未來展望 90
參考文獻 92
自傳 94

[1]吳政忠， “東北季風時期台灣海峽風力預報之研究”， 碩士論文，國防大學理工學院應用物理研究所，桃園，100頁，2004。
[2]蔡清彥、柯文雄、許武榮，數值天氣預報，聯經書局，臺北，第1-2頁，1990。
[3]劉復誠、華文逵、鄭月娥， “日本氣象廳8L-NHM及10L-FLM數值預報模式之初步校驗” 氣象學報，第30卷，第2期，第8-47頁，1984。
[4]Haltiner, G. J., and Williams, R. T., “Some Recent Advances in Numerical Weather Prediction,” Mon. Wea. Rev., Vol. 103, pp.571-590, 1975。
[5]方力脩、陳慧貞， “數值天氣預測之校驗方法綜評”， 氣象學報，第35卷，第1期，第48-62頁，1989。
[6]呂芳川、廖杞昌， “中尺度波譜模式對台灣區域環流預報之研究” ，中正嶺學報，第33卷，第2期，第75-94頁，2005。
[7]吳啟雄、廖杞昌， “2006年6月9日梅雨期豪雨個案模擬研究”， 氣象分析與預報，第197期，第11-22頁，2008。
[8]Orszag, S. A., “Transform method for The Calculation of Vector-Coupled Sums: Application to the Spectral Form of the Vorticity Equation,” J. Atmos. Sci., Vol. 27, No. 6, pp. 890-895, 1970.
[9]Jurraud, M., and Campana, K. A., “Higher Order Finite-Differencing Experiments with a Semi-Implicit Model at the National Meteorological Center,” Mon. Wea. Rev., Vol. 107, No. 4, pp.363-376, 1979。
[10]Tatsumi, Y., “A spectral limited-area model with time dependent lateral boundary conditions and its application to a multi-level primitive equation model,” J. Meteor. Soc. Japan, 64, 637-663, 1986.
[11]Hoyer, J. M., “The ECMWF spectral limited area model,” ECMWF Workshop Proc. on Techniques for Horizontal Discretization in Numerical Weather Prediction Model. 343-359, 1987.
[12]Juang, H.-M.H., and Kanamitsu, M., “The NMC Nested Regional Spectral Model,” Mon. Wea. Rev., Vol. 122, pp.3-26, 1994.
[13]Juang, H.-M.H., Hong, S.Y., and Kanamitsu, M., “The NMC Nested Regional Spectral Model An update,” Bull. Amer. Meteor. Soc., Vol. 78, pp.2125-2143, 1997.
[14]Wang, J. J., “Application of the NCEP Regional Spectral Model to Improve Mesoscale Weather Forecasts in Hawaii,” Weather and Forecasting, Vol. 13, pp. 560-575, 1997.
[15]Hong, S. Y., Juang, H. –M. H., and Zhao, Q., “Implementation of Prognostic Cloud Scheme for a Regional Spectral Model,” Resubmitted to Mon. Wea. Rev., 1997.
[16]Lu, F.-C., Juang, H.-M. H., and Liao, C.-C., “A Numerical Case Study of the Passage of a Cold Surge Across Taiwan,” Meteorol. Atmos. Phys., 95, 27-52, 2007.
[17]Lu, F.-C., Juang, H.-M. H., and Liao, C.-C., “Revisiting Horizontal Diffusion of Perturbations over Terrain for NCEP RSM,” Terr. Atmos. Ocean. Sci., Vol. 18, No.1, pp.67-83, 2007.
[18]任亦偉， “巢狀模式在飛航預報之策略研究”， 碩士論文，國防大學理工學院應用物理研究所，桃園，107頁，2003。