過往研究指出低層水氣資訊與對流系統的發生與否有很高的相關性，而雷達折射率觀測資料對近地表的水氣變化非常敏感，因此可使用雷達低仰角掃描所觀測之折射指數作為低層水氣資訊的依據。本研究中使用系集資料同化系統探討同化雷達回波、徑向風和折射率資料的效果，並選擇於2012年6月11日為北臺灣帶來豐沛的降水，導致各地出現嚴重的淹水災情的梅雨鋒面系統作為研究個案。利用觀測系統模擬實驗進行一系列的同化測試，測試的項目包括：(1)將近地表(~50m)的折射率資訊錯位至0.5度仰角上進行同化，比較其和同化在正確位置上(近地表)的差異；(2)增加水氣局地化半徑的影響；(3)新增溫度場資料同化的效益；(4)檢視新增新竹S-PolKa雷達所觀測之回波、徑向風、折射率資料所帶來的效益；(5)比較在不同模式解析度(3公里與1公里)下同化雷達折射率的差異。 結果顯示同化雷達回波與徑向風資料能有效修正鋒面系統的位置，進一步同化雷達折射率資料能改善近地表濕度場，並提升局部地區的短期預報能力；若將近地表的折射率資料置於PPI上進行同化會產生錯誤的修正；增加水氣的局地化半徑可以修正更大範圍的水氣；同化溫度資料可以藉由溫度與濕度的交相關性修正濕度場，改善降雨預報；位於新竹的S-PolKa雷達海拔高度較低、位置較靠近鋒面帶，能提供更多鋒面底層以及上游的資料，對於重建以及維持鋒面結構有很大的幫助；提升模式解析度能模擬更小尺度的渦流系統，有助於增強折射指數的同化效益。 ;Past studies show a strong link between low-level moisture information and convective initiation(CI). Radar refractivity data is sensitive to the change of near-surface moisture, so refractivity observed by radar low-elevation angle scan can be used as the information about low-level water vapor. In this study, ensemble data assimilation system was used to discuss the effect of assimilating radar reflectivity, radio wind and refractivity. 11 June 2012 Mei-Yu front system case was chosen for research. It brought heavy precipitation over the north western Taiwan, and cause flooding. OSSEs is utilized to conduct a series of sensitivity test, including: (1) Assimilating refractivity data from near surface at 0.5 elevation angle, and compare it with assimilating at right position(near-surface). (2) Impact of increasing localization distance. (3) Improvement of assimilating temperature data. (4) Benefit of additional radar data (reflectivity, radio wind and refractivity) observed by S-PolKa radar. (5) Compare the result between assimilating refractivity data at difference model resolution (3km and 1km). Result shows that assimilating reflectivity and radial wind can correct the position of frontal system. Assimilating radar refractivity can modify near-surface moisture field. Assimilating refractivity at PPI plane will cause wrong adjustment. Increasing localization distance of water vapor field can modify larger area of water vapor field. Assimilating temperature data can modify moisture field by the cross-correlation between temperature and water vapor, and improve precipitation forecast. Since the altitude of S-PolKa radar is low, it can provide more information about the structure of frontal system at lower level. Increasing the resolution of model can simulate smaller scale turbulent convective features, and improve the benefit of assimilating refractivity.
