在0616的個案裡，由於在台灣東部非雷達觀測區域的初始擾動偏濕，而使得實驗結果在此區域有較大的濕偏差。若直接使用ECMWF再分析資料所做的單一預報此偏差情形並沒有那麼嚴重，比對再分析資料後發現，水相粒子在擾動後起轉的分布情形與再分析場的濕度有關，由於此個案的再分析場濕度較大，濕偏差的情形也較嚴重，而此現象也因無觀測而有誤差持續累積的情形。但台灣西南方的降雨一樣有較好的估計，因此綜合兩個個案的實驗結果，使用WRF-LETKF雷達資料同化系統能有效改善梅雨的定量降雨預報結果。 ;The Local Ensemble Transform Kalman Filter (LETKF) method, coupled with Weather Research and Forecast (WRF) model, is applied to assimilate data from five Doppler radars in Taiwan, with the purpose of investigating the improvement on short-term quantitative precipitation forecast (QPF) for rainfall events occurred during the Mei-Yu season. Two heavy precipitation cases from the 2008 SoWMEX IOP#8 field experiments are selected.
The overall results demonstrate that by using WRF-LETKF to assimilate the radar data, the performance of model QPF for representing the Mei-Yu rainfall can be significantly improved. In the first case of June 14, 2008, it is found that by assimilating the 0 dBZ data, the spurious convection can be effectively suppressed. Extending the length of the radar data assimilation to two hours produces better rainfall forecast results. Generating initial perturbations from randomly selected, 6-hr apart data from the NCEP 1ox1o re-analysis data turns out to be a better way to capture the uncertainty related to the Mei-Yu frontal flow than the original NCEP NMC method does.
The same model setup and assimilation method is applied to the second event on June 16, 2008. The pattern and amount of the forecasted rainfall pattern and over southwestern Taiwan indicates a very encouraging result. However, the rainfall prediction over eastern Taiwan becomes unrealistic strong, and this over-estimation cannot be mitigated due to the lack of radar data in this area. This indicates the importance of having a complete radar coverage over Taiwan and vicinity area.
