利用GPS的掩星觀測技術，可反演出高解析度大氣參數的變化，以利於全球大氣的即時觀測。去(2006)年4月15日福衛三號發射升空，因此本文使用CWB/GFS-3DVAR全球同化系統2006年6-8月有、無加入福衛三號掩星的資料，來分析GPS掩星資料對全球氣候研究的影響，並利用NCEP分析資料作一比對。 比對有、無加入GPS資料的結果，其環流結構與大氣駐波的特徵均有不錯的表現，兩者的差異分布主要集中在熱帶、副熱帶，且以海洋上較顯著且有組織性。而有加入GPS的貢獻量約佔各參數的10 % 以下，以海洋居多的南半球改變程度較為明顯，顯示福衛三號資料已有某種程度的影響，特別有助於增加海面上的資訊。最後和NCEP資料比對，熱帶地區的差異較為顯著且分佈於海洋上，其原因可能與NCEP導入較多非傳統觀測資料，尤其在缺乏觀測的海洋所致。 在緯向平均有、無加入GPS資料的分析，8月較6-7月差異量大，原因是7月底前掩星訊號接收機的軟體有問題，直到7月底才修正了軟體所致。兩者的變異數比同樣在南半球的變化較大，原因除了和GPS觀測可增加海面上的資訊外，另一原因可能與南半球為冬季有關。各項參數變化明顯，加入GPS資料對於變化明顯的地方影響較大。然而結構上的差異，各項參數在850mb存在一明顯界線，上下層的結構呈現反相位。與NCEP分析資料比對也有相同的情形，顯示CWB全球同化系統對邊界層資料的處理須再作進一步的探討。 最後藉由水氣傳輸和可降水量的分析， GPS資料加入的貢獻雖與原環流架構無一致對應，但就分布狀況而言雖零散卻仍具有組織性。分析結果顯示差異較大的區域仍集中在熱帶、副熱帶地區，又以海面上的變化最為顯著。同樣對海洋居多的南半球影響較大，顯示福衛三號的加入對全球水文循環也有某種程度的貢獻。 By using the radio occultation (RO) technique, high resolution atmospheric sounding variables can be retrieved from the very accurate Global Positioning System (GPS) measurement and eventually provide a real time climate monitoring. After FORMOSA-3 launched on April 15, 2006, a sequential assimilated data generated by CWB/GFS-3DVAR global assimilation system are utilized to investigate the impact of GPS RO data on the climate study and analysis. To examine the advantage of using GPS RO data, NCEP analysis data are also adapted to serve a cross comparison purpose. The basic atmospheric circulation structure and standing waves are properly captured by CWB/GFS-3DVAR assimilation system with and excluding GPS RO data. However, the major differences appear in both tropics and subtropical regions, especially over the ocean. The estimated statics in terms of variance ratio for each atmospheric variable, in general, are less than 10 % with comparison between with and excluding GPS RO results. Furthermore, the variation shows more significant over the ocean in Southern than Northern hemisphere indicates that GPS RO data have some sort of impact on the climate study. Due to the GPS receiver firmware upgraded for fixing the data retri- eval problem in late July, the latitude-height zonal cross section for each atmospheric variable shows significant variation in August then in June and July. Again, the prominent differences appear in the vast ocean occu- pied Southern hemisphere with GPS RO data. Another possible reason to cause relatively significant variation in Southern hemisphere might be contributed from the mid-latitude cyclone activities during Southern win- ter time. A further examination is along the line on the process. Finally, the primary results from the hydrological analysis also demonstrate that GPS RO data have some influences on global water vapor transport and budget analysis. It also shows that the major impact mainly appear in the tropic and subtropical ocean area.
