本研究旨在探討同化福爾摩沙衛星七號（FORMOSAT-7/COSMIC-2）全球 導航衛星系統掩星觀測（GNSS RO）對梅雨鋒面個案降雨預報的影響。以 2022 年6月6–7日發生於臺灣的梅雨鋒面豪雨個案（TAHOPE IOP3）為研究 對象，使用WRFDA混合資料同化系統（3DEnVar）進行同化實驗，同化 FORMOSAT-7 GNSS RO 資料及全球電信系統（GTS）常規觀測，並使用ERA5再分析資料進行預報效果驗證。 結果顯示，同化GNSS RO資料（EPH組）顯著改善模式初始分析場的水 汽與溫度結構，特別是在鋒面附近增加南側海洋上低層水汽含量，降低鋒面附近的溫度場誤差，改善了鋒面的結構特徵，使分析場更貼近ERA5再分析資料。此外，透過模式對不同降雨時段的詳細暴雨成因分析可知，同化GNSS RO資料可更準確捕捉鋒面附近的水汽輻合與對流發展過程，有效提升降水位置及強度的預報表現。同時，透過系集敏感度分析顯示，GNSS RO資料的加入有助於降低系集預報的不確定性，增強系集成員間降雨預報的一致性，尤其對短期（0–24小時）降雨預報的改善最為顯著。整體而言，EPH組在各個預報時段內的均方根誤差（RMSE）均較僅同化GTS組顯著降低，空間相關係數(SCC）則顯著提升。 綜合而言，本研究證實同化FORMOSAT-7 GNSS RO資料能有效提升梅雨 鋒面豪雨個案的預報性能，尤其在暴雨成因分析與系集預報穩定性方面的提升更具實際預報價值。;This study examines the impact of assimilating Global Navigation Satellite System Radio Occultation (GNSS RO) observations on the rainfall forecast of a Mei-Yu frontal heavy precipitation event that occurred on 6–7 June 2022 (TAHOPE IOP3). The Weather Research and Forecasting (WRF) model, coupled with the hybrid three dimensional ensemble-variational (3DEnVar) data assimilation system (WRFDA), was employed to incorporate both GNSS RO and conventional observations into the model’s initial conditions.
To evaluate the influence of GNSS RO assimilation, a simulation experiment that assimilated both GNSS RO and conventional data (referred to as EPH) was compared with a control experiment that assimilated only conventional observations (referred to as GTS). The results demonstrated that GNSS RO assimilation notably improved the moisture and temperature distributions near the Mei-Yu front, yielding a frontal structure more consistent with the ERA5 reanalysis.
Compared to the GTS experiment, the EPH simulation exhibited lower forecast errors and higher correlation coefficients for temperature, moisture, and wind fields. In terms of precipitation, the EPH experiment more accurately reproduced both the spatial distribution and intensity of the heavy rainfall along the front. Forecast verification metrics further confirmed that the EPH outperformed the GTS run.
In summary, assimilating GNSS RO observations significantly enhanced the model analysis and precipitation forecast for this Mei-Yu frontal heavy rainfall event.
