本計畫上一年度已順利在GSI系統下加入系集預報成員以建立混成(Hybrid)同化系統之初步雛型。由此初步之混成(Hybrid)同化系統所得的分析場顯示南半球預報有顯著的改善。雖然系集預報成員是以簡單的時間延遲方法所產生，但此結果已充分的顯示出Hybrid系統對於改善作業分析與預報的潛力。今年度的整體目標將著重在引進NCEP GFS的系集同化系統，以提供更加適當的系集成員為hybrid系統之用，並完成在CWB/GFS架構下的建置與初步測試。另外，此系統的建置也可提供GPSRO觀測另一個更為進階的同化系統，以探討加入流場相依的背景誤差是否可更提高GPSRO觀測對於CWB/GFS的影響。 ; In the last fiscal year of this project, it had been demonstrated that a significant improvement of forecast for Southern-hemisphere is achieved by using the prototype GSI-hybrid system at the CWB. Although the ensemble members were generated from a “lagged-average forecasting” method, result shows the potential advantage of the 3DVAR-EnKF hybrid system on advancing the performance of analysis and forecast of the operational CWB/GFS. The first objective of this project in this year is to implement the ensemble data assimilation system of NCEP GFS under the CWB/GFS’s framework in order to provide appropriate ensemble members for the GSI-hybrid system. In addition, this new advanced data assimilation system provides another alternative to investigate whether the impact of the assimilation of GPSRO can be further improved by incorporating the flow-dependent the information in the background error statistics. ; 研究期間 10201 ~ 10212
